How far have we come? Artificial intelligence for chest radiograph interpretation.

Due to recent advances in artificial intelligence, there is renewed interest in automating interpretation of imaging tests. Chest radiographs are particularly interesting due to many factors: relatively inexpensive equipment, importance to public health, commonly performed throughout the world, and deceptively complex taking years to master. This article presents a brief introduction to artificial intelligence, reviews the progress to date in chest radiograph interpretation, and provides a snapshot of the available datasets and algorithms available to chest radiograph researchers. Finally, the limitations of artificial intelligence with respect to interpretation of imaging studies are discussed.

A Query Expansion Framework in Image Retrieval Domain Based on Local and Global Analysis.

We present an image retrieval framework based on automatic query expansion in a concept feature space by generalizing the vector space model of information retrieval. In this framework, images are represented by vectors of weighted concepts similar to the keyword-based representation used in text retrieval. To generate the concept vocabularies, a statistical model is built by utilizing Support Vector Machine (SVM)-based classification techniques. The images are represented as "bag of concepts" that comprise perceptually and/or semantically distinguishable color and texture patches from local image regions in a multi-dimensional feature space. To explore the correlation between the concepts and overcome the assumption of feature independence in this model, we propose query expansion techniques in the image domain from a new perspective based on both local and global analysis. For the local analysis, the correlations between the concepts based on the co-occurrence pattern, and the metrical constraints based on the neighborhood proximity between the concepts in encoded images, are analyzed by considering local feedback information. We also analyze the concept similarities in the collection as a whole in the form of a similarity thesaurus and propose an efficient query expansion based on the global analysis. The experimental results on a photographic collection of natural scenes and a biomedical database of different imaging modalities demonstrate the effectiveness of the proposed framework in terms of precision and recall.

Content-based image retrieval for scientific literature access.

OBJECTIVES: An increasing number of articles are published electronically in the scientific literature, but access is limited to alphanumerical search on title, author, or abstract, and may disregard numerous figures. In this paper, we estimate the benefits of using content-based image retrieval (CBIR) on article figures to augment traditional access to articles. METHODS: We selected four high-impact journals from the Journal Citations Report (JCR) 2005. Figures were automatically extracted from the PDF article files, and manually classified on their content and number of sub-figure panels. We make a quantitative estimate by projecting from data from the Cross-Language Evaluation Forum (ImageCLEF) campaigns, and qualitatively validate it through experiments using the Image Retrieval in Medical Applications (IRMA) project. RESULTS: Based on 2077 articles with 11,753 pages, 4493 figures, and 11,238 individual images, the predicted accuracy for article retrieval may reach 97.08%. CONCLUSIONS: Therefore, CBIR potentially has a high impact in medical literature search and retrieval.

Assessment of an ensemble of machine learning models toward abnormality detection in chest radiographs.

Respiratory diseases account for a significant proportion of deaths and disabilities across the world. Chest X-ray (CXR) analysis remains a common diagnostic imaging modality for confirming intra-thoracic cardiopulmonary abnormalities. However, there remains an acute shortage of expert radiologists, particularly in under-resourced settings, resulting in severe interpretation delays. These issues can be mitigated by a computer-aided diagnostic (CADx) system to supplement decision-making and improve throughput while preserving and possibly improving the standard-of-care. Systems reported in the literature or popular media use handcrafted features and/or data-driven algorithms like deep learning (DL) to learn underlying data distributions. The remarkable success of convolutional neural networks (CNN) toward image recognition tasks has made them a promising choice for automated medical image analyses. However, CNNs suffer from high variance and may overfit due to their sensitivity to training data fluctuations. Ensemble learning helps to reduce this variance by combining predictions of multiple learning algorithms to construct complex, non-linear functions and improve robustness and generalization. This study aims to construct and assess the performance of an ensemble of machine learning (ML) models applied to the challenge of classifying normal and abnormal CXRs and significantly reducing the diagnostic load of radiologists and primary-care physicians.

A spine X-ray image retrieval system using partial shape matching.

In recent years, there has been a rapid increase in the size and number of medical image collections. Thus, the development of appropriate methods for medical information retrieval is especially important. In a large collection of spine X-ray images, maintained by the National Library of Medicine, vertebral boundary shape has been determined to be relevant to pathology of interest. This paper presents an innovative partial shape matching (PSM) technique using dynamic programming (DP) for the retrieval of spine X-ray images. The improved version of this technique called corner-guided DP is introduced. It uses nine landmark boundary points for DP search and improves matching speed by approximately 10 times compared to traditional DP. The retrieval accuracy and processing speed of the retrieval system based on the new corner-guided PSM method are evaluated and included in this paper.

A novel stacked generalization of models for improved TB detection in chest radiographs.

Chest x-ray (CXR) analysis is a common part of the protocol for confirming active pulmonary Tuberculosis (TB). However, many TB endemic regions are severely resource constrained in radiological services impairing timely detection and treatment. Computer-aided diagnosis (CADx) tools can supplement decision-making while simultaneously addressing the gap in expert radiological interpretation during mobile field screening. These tools use hand-engineered and/or convolutional neural networks (CNN) computed image features. CNN, a class of deep learning (DL) models, has gained research prominence in visual recognition. It has been shown that Ensemble learning has an inherent advantage of constructing non-linear decision making functions and improve visual recognition. We create a stacking of classifiers with hand-engineered and CNN features toward improving TB detection in CXRs. The results obtained are highly promising and superior to the state-of-the-art.

Patterns of neuropsychological impairment in frontotemporal dementia.

OBJECTIVE: To differentiate frontotemporal dementia (FTD) subtypes from each other and from probable Alzheimer disease (AD) using neuropsychological tests. METHODS: Patients with FTD and AD (n = 109) were studied with a comprehensive neuropsychological protocol at first contact. Data were subjected to a principal components analysis (PCA) to extract core neuropsychological features. A five-factor solution accounted for 72.89% of the variance and yielded factors related to declarative memory, working memory/visuoconstruction, processing speed/mental flexibility, lexical retrieval, and semantic memory. RESULTS: Between- and within-group analyses revealed that patients with AD obtain their lowest scores on tests of declarative memory while semantic dementia (SemD) patients are particularly disadvantaged on tests of semantic memory. On tests of processing speed/mental flexibility time to completion was faster for social comportment/dysexecutive (SOC/EXEC) patients, but these patients made more errors on some tests. Patients with corticobasal degeneration (CBD) and progressive nonfluent aphasia (PNFA) were impaired on tests of working memory. Logistic regression analyses using factor scores successfully assigned FTD subgroups and AD patients into their respective diagnostic categories. CONCLUSION: Patients with differing frontotemporal dementia phenotypes can be distinguished from each other and from Alzheimer disease using neuropsychological tests.

Trying to tell a tale: discourse impairments in progressive aphasia and frontotemporal dementia.

OBJECTIVE: To assess discourse in patients with frontotemporal dementia (FTD). METHODS: The authors asked patients with progressive nonfluent aphasia (PNFA), patients with semantic dementia (SemD), and nonaphasic patients with a disorder of social comportment and executive functioning (SOC/EXEC) to narrate the story of a wordless children's picture book. RESULTS: The authors found significant discourse impairments in all three groups of patients. Moreover, there were qualitatively important differences between the groups. Patients with PNFA had the sparsest output, producing narratives with the fewest words per minute. Patients with SemD had difficulty retrieving words needed to tell their narratives. Though not aphasic, patients with SOC/EXEC had profound difficulty organizing their narratives, and they could not effectively express the point of the story. This deficit correlated with poor performance on a measure of executive resources requiring an organized mental search. In addition, a correlation of narrative organization with cortical atrophy in patients with SOC/EXEC was significant in right frontal and anterior temporal brain regions. CONCLUSIONS: Impaired day-to-day communication in nonaphasic frontotemporal dementia patients with a disorder of social comportment and executive functioning is due in part to a striking deficit in discourse organization associated with right frontotemporal disease. Difficulty with discourse in progressive aphasia is due largely to the language impairments of these patients.

Metacognitive deficits in frontotemporal dementia.

OBJECTIVES: To investigate whether metacognitive impairments in self-awareness and self-monitoring occur in patients with frontotemporal dementia (FTD), particularly among those with prominent social and dysexecutive impairments. METHODS: Patients diagnosed with FTD were divided by clinical subtype (social-dysexecutive (n = 12) aphasic (n = 15), and constituent subgroups of progressive non-fluent aphasia and semantic dementia) and compared with subjects with probable Alzheimer's disease (AD, n = 11) and age-matched healthy controls (n = 11). All subjects completed comprehensive behavioural ratings scales, which were compared with caregiver ratings. Subjects also rated their test performances in verbal associative fluency, word list learning, and memory task with comparisons made between actual and judged performance levels. RESULTS: The FTD sample as a whole showed significantly less behavioural self-awareness and self-knowledge than the AD and healthy control samples. FTD patients with prominent social and dysexecutive impairments demonstrated the most extensive loss of self-awareness and self-knowledge, significantly overrating themselves in multiple social, emotional, and cognitive domains, and failing to acknowledge that any behavioural change had occurred in most areas. The remaining clinical samples showed select and minimal discrepancies. All clinical groups were significantly unaware of their apathy levels. Most FTD patients judged episodic cognitive test performance adequately, with partial difficulties observed in the socially impaired and progressive non-fluent aphasia subgroups. CONCLUSIONS: FTD patients, particularly those with prominent social and dysexecutive impairments, exhibit profound metacognitive anosognosia that may represent a loss of self-awareness, self-monitoring, and self-knowledge, likely related to significant prefrontal pathophysiology. Other FTD clinical groups and AD patients showed less pervasive and more select metacognitive deficiencies.