Enhancement of Neurocognitive Assessments Using Smartphone Capabilities: Systematic Review

Abstract

Background Comprehensive exams such as the Dean-Woodcock Neuropsychological Assessment System, the Global Deterioration Scale, and the Boston Diagnostic Aphasia Examination are the gold standard for doctors and clinicians in the preliminary assessment and monitoring of neurocognitive function in conditions such as neurodegenerative diseases and acquired brain injuries (ABIs). In recent years, there has been an increased focus on implementing these exams on mobile devices to benefit from their configurable built-in sensors, in addition to scoring, interpretation, and storage capabilities. As smartphones become more accepted in health care among both users and clinicians, the ability to use device information (eg, device position, screen interactions, and app usage) for subject monitoring also increases. Sensor-based assessments (eg, functional gait using a mobile device’s accelerometer and/or gyroscope or collection of speech samples using recordings from the device’s microphone) include the potential for enhanced information for diagnoses of neurological conditions; mapping the development of these conditions over time; and monitoring efficient, evidence-based rehabilitation programs. Objective This paper provides an overview of neurocognitive conditions and relevant functions of interest, analysis of recent results using smartphone and/or tablet built-in sensor information for the assessment of these different neurocognitive conditions, and how human-device interactions and the assessment and monitoring of these neurocognitive functions can be enhanced for both the patient and health care provider. Methods This survey presents a review of current mobile technological capabilities to enhance the assessment of various neurocognitive conditions, including both neurodegenerative diseases and ABIs. It explores how device features can be configured for assessments as well as the enhanced capability and data monitoring that will arise due to the addition of these features. It also recognizes the challenges that will be apparent with the transfer of these current assessments to mobile devices. Results Built-in sensor information on mobile devices is found to provide information that can enhance neurocognitive assessment and monitoring across all functional categories. Configurations of positional sensors (eg, accelerometer, gyroscope, and GPS), media sensors (eg, microphone and camera), inherent sensors (eg, device timer), and participatory user-device interactions (eg, screen interactions, metadata input, app usage, and device lock and unlock) are all helpful for assessing these functions for the purposes of training, monitoring, diagnosis, or rehabilitation. Conclusions This survey discusses some of the many opportunities and challenges of implementing configured built-in sensors on mobile devices to enhance assessments and monitoring of neurocognitive functions as well as disease progression across neurodegenerative and acquired neurological conditions.

Design of a neurocognitive digital health system (NDHS) for neurodegenerative diseases

“…As digital health technology is becoming more commonplace in the healthcare arena, integration of device sensors and capabilities for the collection of relevant and objective data increases [47]. Previous work has identified sensor-based digital assessments (e.g., accelerometry based gait assessments or speech recognition systems for healthcare) which provide promising user-device interactions for the collection of objective digital biomarkers across functional areas of neurocognition [5,17,28,41,50,55].…”

“…Using device-based sensors and/or interactions in the formation and configuration of functional tasks enhances the utility and quality of collected data. Additionally, with increased opportunity for user participation on their own devices and the ability of clinicians to collect and analyze enhanced objective datasets, digital health systems become a robust modality for the administration of neurocognitive assessments [47].…”

“…Given different devices and collection methodologies, both objective (e.g., clinical diagnoses, numerical physiologic data, neurocognitive assessment scores) and subjective (e.g., patient or caretaker reported outcomes) information is collected. This objective and subjective data is necessary for disease classification in addition to the comprehensive monitoring and rehabilitation of said condition [25,30,47]. The information collected from each NDHS component, across functional areas of motor (e.g., fine and gross motor), memory (e.g., long and short term), speech (e.g., frequency, variations, and repeatability), executive function (e.g., judgement and planning), sensory (e.g., visual, tactile, and aural), language (e.g., semantics, syntax, and pragmatics), behavioral and psychological function (e.g., emotion), sleep (e.g., quality and duration), and autonomic function (e.g., heart rate, blood oxygen saturation), are seen in Table 1.…”

“…Traditionally, neurocognitive functions of interest have been assessed in clinical settings using various accepted pen-and-paper style assessments (e.g., Montreal Cognitive Assessment (MoCA) [31], Mini Mental State Examination (MMSE) [49], and the Menu Task Assessment (MT) [1]) [52]. However, these assessments only evaluate a subset of relevant neurocognitive functions [47]. Neurocognitive functions of motor, memory, speech, language, executive function, sensory, behavioral and psychological function, sleep, and autonomic functions should all be evaluated as each of them may be difficult for individuals with neurodegenerative diseases [3,7,9,11,15,21,23,29,33].…”

“…Neurocognitive functions of motor, memory, speech, language, executive function, sensory, behavioral and psychological function, sleep, and autonomic functions should all be evaluated as each of them may be difficult for individuals with neurodegenerative diseases [3,7,9,11,15,21,23,29,33]. The use of digital health technology and its capabilities (e.g., device configurations, inherent device sensors, and human-device interactions) allows for the collection of far more information and objective metrics than we ever could achieve using pen-and-paper style tests, while also assessing all functional areas of neurocognition [10,47]. Further, digital health systems can also support continuous and objective health monitoring, encourage healthy behavior, support chronic disease self-management, enhance clinician knowledge, and aid in the personalization of interventions [20].…”