Pseudo-class part prototype networks for interpretable breast cancer classification.

Interpretability in machine learning has become increasingly important as machine learning is being used in more and more applications, including those with high-stakes consequences such as healthcare where Interpretability has been regarded as a key to the successful adoption of machine learning models. However, using confounding/irrelevant information in making predictions by deep learning models, even the interpretable ones, poses critical challenges to their clinical acceptance. That has recently drawn researchers' attention to issues beyond the mere interpretation of deep learning models. In this paper, we first investigate application of an inherently interpretable prototype-based architecture, known as ProtoPNet, for breast cancer classification in digital pathology and highlight its shortcomings in this application. Then, we propose a new method that uses more medically relevant information and makes more accurate and interpretable predictions. Our method leverages the clustering concept and implicitly increases the number of classes in the training dataset. The proposed method learns more relevant prototypes without any pixel-level annotated data. To have a more holistic assessment, in addition to classification accuracy, we define a new metric for assessing the degree of interpretability based on the comments of a group of skilled pathologists. Experimental results on the BreakHis dataset show that the proposed method effectively improves the classification accuracy and interpretability by respectively 8 % and 18 % . Therefore, the proposed method can be seen as a step toward implementing interpretable deep learning models for the detection of breast cancer using histopathology images.

On the interpretability of part-prototype based classifiers: a human centric analysis.

Part-prototype networks have recently become methods of interest as an interpretable alternative to many of the current black-box image classifiers. However, the interpretability of these methods from the perspective of human users has not been sufficiently explored. In addition, previous works have had major issues with following proper experiment design and task representation that limit their reliability and validity. In this work, we have devised a framework for evaluating the interpretability of part-prototype-based models from a human perspective that solves these issues. The proposed framework consists of three actionable metrics and experiments. The results of these experiments will reveal important and reliable interpretability related properties of such models. To demonstrate the usefulness of our framework, we performed an extensive set of experiments using Amazon Mechanical Turk. They not only show the capability of our framework in assessing the interpretability of various part-prototype-based models, but they also are, to the best of our knowledge, the most comprehensive work on evaluating such methods in a unified framework.

Predictive modelling of Parkinson's disease progression based on RNA-Sequence with densely connected deep recurrent neural networks.

The advent of recent high throughput sequencing technologies resulted in unexplored big data of genomics and transcriptomics that might help to answer various research questions in Parkinson's disease (PD) progression. While the literature has revealed various predictive models that use longitudinal clinical data for disease progression, there is no predictive model based on RNA-Sequence data of PD patients. This study investigates how to predict the PD Progression for a patient's next medical visit by capturing longitudinal temporal patterns in the RNA-Seq data. Data provided by Parkinson Progression Marker Initiative (PPMI) includes 423 PD patients without revealing any race, sex, or age information with a variable number of visits and 34,682 predictor variables for 4 years. We propose a predictive model based on deep Recurrent Neural Network (RNN) with the addition of dense connections and batch normalization into RNN layers. The results show that the proposed architecture can predict PD progression from high dimensional RNA-seq data with a Root Mean Square Error (RMSE) of 6.0 and a rank-order correlation of (r = 0.83, p < 0.0001) between the predicted and actual disease status of PD.

Multiview Feature Selection for Single-View Classification.

In many real-world scenarios, data from multiple modalities (sources) are collected during a development phase. Such data are referred to as multiview data. While additional information from multiple views often improves the performance, collecting data from such additional views during the testing phase may not be desired due to the high costs associated with measuring such views or, unavailability of such additional views. Therefore, in many applications, despite having a multiview training data set, it is desired to do performance testing using data from only one view. In this paper, we present a multiview feature selection method that leverages the knowledge of all views and use it to guide the feature selection process in an individual view. We realize this via a multiview feature weighting scheme such that the local margins of samples in each view are maximized and similarities of samples to some reference points in different views are preserved. Also, the proposed formulation can be used for cross-view matching when the view-specific feature weights are pre-computed on an auxiliary data set. Promising results have been achieved on nine real-world data sets as well as three biometric recognition applications. On average, the proposed feature selection method has improved the classification error rate by 31 percent of the error rate of the state-of-the-art.

Talk2Me: Automated linguistic data collection for personal assessment.

Language is one the earliest capacities affected by cognitive change. To monitor that change longitudinally, we have developed a web portal for remote linguistic data acquisition, called Talk2Me, consisting of a variety of tasks. In order to facilitate research in different aspects of language, we provide baselines including the relations between different scoring functions within and across tasks. These data can be used to augment studies that require a normative model; for example, we provide baseline classification results in identifying dementia. These data are released publicly along with a comprehensive open-source package for extracting approximately two thousand lexico-syntactic, acoustic, and semantic features. This package can be applied arbitrarily to studies that include linguistic data. To our knowledge, this is the most comprehensive publicly available software for extracting linguistic features. The software includes scoring functions for different tasks.

A Machine Learning Framework for Automatic and Continuous MMN Detection With Preliminary Results for Coma Outcome Prediction.

Mismatch negativity (MMN) is a component of the event-related potential (ERP) that is elicited through an odd-ball paradigm. The existence of the MMN in a coma patient has a good correlation with coma emergence; however, this component can be difficult to detect. Previously, MMN detection was based on visual inspection of the averaged ERPs by a skilled clinician, a process that is expensive and not always feasible in practice. In this paper, we propose a practical machine learning (ML) based approach for detection of MMN component, thus, improving the accuracy of prediction of emergence from coma. Furthermore, the method can operate on an automatic and continuous basis thus alleviating the need for clinician involvement. The proposed method is capable of the MMN detection over intervals as short as two minutes. This finer time resolution enables identification of waxing and waning cycles of a conscious state. An auditory odd-ball paradigm was applied to 22 healthy subjects and 2 coma patients. A coma patient is tested by measuring the similarity of the patient's ERP responses with the aggregate healthy responses. Because the training process for measuring similarity requires only healthy subjects, the complexity and practicality of training procedure of the proposed method are greatly improved relative to training on coma patients directly. Since there are only two coma patients involved with this study, the results are reported on a very preliminary basis. Preliminary results indicate we can detect the MMN component with an accuracy of 92.7% on healthy subjects. The method successfully predicted emergence in both coma patients when conventional methods failed. The proposed method for collecting training data using exclusively healthy subjects is a novel approach that may prove useful in future, unrelated studies where ML methods are used.

Development of a Smart Home-Based Package for Unobtrusive Physiological Monitoring.

Home-based programs have been shown to be effective in improving health conditions, patient self-management, quality of life, and health outcomes. However, there is mixedevidence on the effectiveness of these programs due to limitations in the intervention tools that are used, primarily the burden that is placed on the user, especially among seniors. In this paper we developed a novel home-based package that measures critically important physiological information such that neither active compliance or interaction from the user is required. To this end, we embedded passive sensors (including load cells, electrodes, pulse sensor and color sensors) into common household items such as tiling, furniture and wall. The smart package measures subject's electrocardiogram (ECG), photoplethysmogram (PPG), ballistocardiogram (BCG), electromyogram (EMG) and imaging photoplethysmogram (IPPG). In contrast to the previous studies, the proposed package measures all the physiological information unobtrusively, simultaneously and in a synchronized manner such that all the data samples corresponding to different intervals of a specific cardiovascular cycle can be identified. Such information can be analyzed by a clinician or be used for a higher level information extraction such as beat-to-beat blood pressure estimation. In addition, the proposed package is the first and only homebased technology that can simultaneously and unobtrusively capture both mechanical and electrical characteristics of user's heart activities. This results in a more accurate home-based vital parameters monitoring.

40-Hz ASSR for Measuring Depth of Anaesthesia During Induction Phase.

This paper proposes an anaesthesia monitoring system that accurately measures the depth of anaesthesia through 40-Hz auditory steady-state response. With accurate and fast depth of anaesthesia measuring, the monitor can reduce the incidence of awareness during surgical operation. The proposed denoising method for extracting 40-Hz auditory steady-state cycles, adaptive multilevel wavelet denoising, enabled the system to extract auditory steady-state response cycles from fewer epochs and over short periods of time which is of crucial importance in monitoring anaesthesia. The noise estimation scheme, adaptive threshold levels, rearranging, and multilevel denoising of frames increase the accuracy and signal to noise ratio of the extracted cycles. The modified fuzzy c-means clustering scheme, proposed to improve clustering performance in noisy data bases where no prior information about the level of noise and signal energy is available, is used for clustering the auditory steady-state cycles. Weighting the features with a novel algorithm and based on their differentiating role in clustering, the modified fuzzy c-means improves fuzziness in cluster partitions and the geometrical structure of the data. An index called depth of anaesthesia index is defined and determined at each cycle based on the clustering information of the cycle and the previous ones. The algorithm is applied to auditory steady-state response signals recorded from 20 human subjects during surgical operations with Propofol-induced general anaesthesia. The accuracy of the depth of anaesthesia index is validated through the subjects' medical markers, clinical parameters, and the recorded bispectral index during the induction phase. Depth of anaesthesia index is verified to be accurate and able to detect fast transitions between different levels of anaesthesia. The computed depth of anaesthesia indices detected the induction of anaesthesia on average 55 s faster than bispectral index and 17 s earlier than loss of eyelash reflex.

Logistic Localized Modeling of the Sample Space for Feature Selection and Classification.

Conventional feature selection algorithms assign a single common feature set to all regions of the sample space. In contrast, this paper proposes a novel algorithm for localized feature selection for which each region of the sample space is characterized by its individual distinct feature subset that may vary in size and membership. This approach can therefore select an optimal feature subset that adapts to local variations of the sample space, and hence offer the potential for improved performance. Feature subsets are computed by choosing an optimal coordinate space so that, within a localized region, within-class distances and between-class distances are, respectively, minimized and maximized. Distances are measured using a logistic function metric within the corresponding region. This enables the optimization process to focus on a localized region within the sample space. A local classification approach is utilized for measuring the similarity of a new input data point to each class. The proposed logistic localized feature selection (lLFS) algorithm is invariant to the underlying probability distribution of the data; hence, it is appropriate when the data are distributed on a nonlinear or disjoint manifold. lLFS is efficiently formulated as a joint convex/increasing quasi-convex optimization problem with a unique global optimum point. The method is most applicable when the number of available training samples is small. The performance of the proposed localized method is successfully demonstrated on a large variety of data sets. We demonstrate that the number of features selected by the lLFS method saturates at the number of available discriminative features. In addition, we have shown that the Vapnik-Chervonenkis dimension of the localized classifier is finite. Both these factors suggest that the lLFS method is insensitive to the overfitting issue, relative to other methods.

Feature Selection for Nonstationary Data: Application to Human Recognition Using Medical Biometrics.

Electrocardiogram (ECG) and transient evoked otoacoustic emission (TEOAE) are among the physiological signals that have attracted significant interest in biometric community due to their inherent robustness to replay and falsification attacks. However, they are time-dependent signals and this makes them hard to deal with in across-session human recognition scenario where only one session is available for enrollment. This paper presents a novel feature selection method to address this issue. It is based on an auxiliary dataset with multiple sessions where it selects a subset of features that are more persistent across different sessions. It uses local information in terms of sample margins while enforcing an across-session measure. This makes it a perfect fit for aforementioned biometric recognition problem. Comprehensive experiments on ECG and TEOAE variability due to time lapse and body posture are done. Performance of the proposed method is compared against seven state-of-the-art feature selection algorithms as well as another six approaches in the area of ECG and TEOAE biometric recognition. Experimental results demonstrate that the proposed method performs noticeably better than other algorithms.

Local Feature Selection for Data Classification.

Typical feature selection methods choose an optimal global feature subset that is applied over all regions of the sample space. In contrast, in this paper we propose a novel localized feature selection (LFS) approach whereby each region of the sample space is associated with its own distinct optimized feature set, which may vary both in membership and size across the sample space. This allows the feature set to optimally adapt to local variations in the sample space. An associated method for measuring the similarities of a query datum to each of the respective classes is also proposed. The proposed method makes no assumptions about the underlying structure of the samples; hence the method is insensitive to the distribution of the data over the sample space. The method is efficiently formulated as a linear programming optimization problem. Furthermore, we demonstrate the method is robust against the over-fitting problem. Experimental results on eleven synthetic and real-world data sets demonstrate the viability of the formulation and the effectiveness of the proposed algorithm. In addition we show several examples where localized feature selection produces better results than a global feature selection method.

Automatic and continuous assessment of ERPs for mismatch negativity detection.

Accurate and fast detection of event related potential (ERP) components is an unresolved issue in neuroscience and critical health care. Mismatch negativity (MMN) is a component of the ERP to an odd stimulus in a sequence of identical stimuli which has good correlation with coma awakening. All of the previous studies for MMN detection are based on visual inspection of the averaged ERPs (over a long recording time) by a skilled neurophysiologist. However, in practical situations, such an expert may not be available or familiar with all aspects of evoked potential methods. Further, we may miss important clinically essential events due to the implicit averaging process used to acquire the ERPs. In this paper we propose a practical machine learning approach for automatic and continuous assessment of the ERPs for detecting the presence of the MMN component. The proposed method is realized in a classification framework. Performance of the proposed method is demonstrated on 22 healthy subjects through a leave-one subject-out strategy where the MMN components are identified with about 93% accuracy.