During the past two decades, innovations in intelligent healthcare systems have revolutionized the ways in which services are provided. These systems are now used in a wide range of settings, from healthcare units to the patient’s home, where they contribute to the efficiency and cost effectiveness of medical procedures or patient support and monitoring. Nonetheless, there are still important challenges in designing and deploying intelligent healthcare systems. The five selected papers of this special issue, written by 16 authors and coauthors from seven countries, provide current knowledge and expertise regarding these challenges for some of today’s most pressing healthcare problems: monitoring and assisted living for the elderly and chronically ill patients, and cancer diagnosis.
The fact that the population is aging in many countries calls for the development of healthcare systems that offer long-term support to individuals with chronic illnesses or cognitive impairments. For example, chronic diseases and aged care accounted for over 70% of healthcare expenditure in Australia, and their cost is projected to increase significantly in the future. In this context, the first paper, “Telehealth monitoring of patients in the community” by Sparks et al. , outlines a decision-support system for the management of chronic illnesses. The system is designed to support community nurses in monitoring the well-being of their patients while reducing the time spent on activities that are either conducted by the system (e.g. measurement of vital signs) or become unnecessary as a result of its use (e.g. driving to and from patients’ homes). This paper particularly illustrates how it is important to match what an intelligent healthcare system offers with the capabilities and needs of its end users. Specifically, a broad range of visualization techniques is discussed (e.g. from trend plots to parallel coordinates) to monitor patients’ signals, and the information provided by these visualizations is related to the level of mathematical capability of the nurses. For example, visualizations of correlations and principle components can be of use for one nurse, while another may prefer a simpler multivariate view of a patient’s well-being. The work of Sparks et al. on chronic diseases is complemented by the second paper , “Ambient assisted living technologies for aging well: a scoping review” by Blackman et al., which focuses on aged care. In this review, 59 technologies were identified from English language peer-reviewed publications published between 2000 and 2013. Technologies that could be used by people living with some degree of cognitive impairment were of particular interest. These technologies were classified and analyzed with respect to emerging themes and solutions in relation to the needs of older adults with cognitive impairments. As the authors pointed out, these needs are multidimensional (e.g. physical and mental), and the population of older adults is heterogeneous; thus, intelligent healthcare systems would benefit from interdisciplinary collaborations to adequately address the variety of needs.
Specific techniques and databases are introduced in the next two articles to support telemonitoring of the elderly and those with impairments. The third paper, “Everyday life sounds database: telemonitoring of elderly or disabled” by Abdoune and Fezari  contributes to the literature on smart homes by constructing and using a database of everyday life sounds to detect activities. The database divides sounds into critical (e.g. broken glass, screaming) and normal sounds, which are themselves subdivided as useful or disturbing (e.g. background noise). These sounds are then used by a system that separates them and uses a classifier to recognize the associated activity, such as screaming or door knocking. The fourth paper, “Employing emotion cues to verify speakers in emotional talking environments” by Shahin  explores another use of sound processing with potential for telemonitoring. Specifically, individuals may be expressing different emotions such as happiness, anger, or sadness as a result of either their health status or what happened in their home. These emotions can make it difficult to be recognized by a speaker verification system, which can be a component of the overall telemonitoring system. Shahin presents an approach that uses classifiers to enhance speaker verification performance in this context, by combining emotion recognizer and speaker recognizer into one recognizer. The approach is evaluated on two databases, showing that the emotional state of the speaker does negatively impact speaker verification performance, but that the proposed approach can better identify the speaker compared to previous studies.
Finally, the fifth article , “Mining breast cancer classification rules from mammograms” by Yeh et al., provides a different application of classifiers from the work of Shahin within the context of intelligent healthcare systems. This work introduces a new methodology for analyzing irregularities in mammograms and creates a set of decision rules that can support radiologists both in classifying abnormal lesions and in classifying tumor severity. Similarly to the article of Sparks et al., these decision rules can support practitioners in conducting repetitive and time-consuming tasks. The decision rules are evaluated on two datasets, where they achieve an average accuracy of 73.64% on the classification of abnormal lesions and 87.14% on the classification of tumor severity.
The guest editors would like to express their gratitude to Hasan Fleyeh, Editor-in-Chief, for his ongoing support for this special issue. We hope that the readers will benefit from the findings and experiences shared in this special issue covering intelligent healthcare systems to assist practitioners in performing diagnoses and monitoring as well as to empower individuals in the management of their well-being.
R. Sparks, B. Celler, C. Okugami, R. Jayasena and M. Varnfield, Telehealth monitoring of patients in the community, J. Intell. Syst. 25 (2016), 37–53.Google Scholar
S. Blackman, C. Matlo, C. Bobrovitskiy, A. Waldoch, M. L. Fang, P. Jackson, A. Mihailidis, L. Nygård, A. Astell and A. Sixsmith, Ambient assisted living technologies for aging well: a scoping review, J. Intell. Syst. 25 (2016), 55–69Google Scholar
L. Abdoune and M. Fezari, Everyday life sounds database: telemonitoring of elderly or disabled, J. Intell. Syst. 25 (2016), 71–84Google Scholar
I. Shahin, Employing emotion cues to verify speakers in emotional talking environments, J. Intell. Syst. 25 (2016), 3–17.Google Scholar
J.-Y. Yeh, S.-W. Chan and T.-H. Wu, Mining breast cancer classification rules from mammograms, J. Intell. Syst. 5 (2016), 19–36.Google Scholar