This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.
Telestroke has developed rapidly as an assessment tool for patients eligible for reperfusion therapy.
To investigate whether vascular neurologists can diagnose intracranial large vessel occlusion (LVO) as quickly and accurately using a smartphone application compared to a hospital-based desktop PC monitor.
We retrospectively enrolled 108 consecutive patients with acute ischemic stroke in the middle cerebral artery territory who underwent magnetic resonance imaging (MRI) within 24 hours of their stroke onset. Two vascular neurologists, blinded to all clinical information, independently evaluated magnetic resonance angiography and fluid-attenuated inversion recovery images for the presence or absence of LVO in the internal carotid artery and middle cerebral artery (M1, M2, or M3) on both a smartphone application (Smartphone-LVO) and a hospital-based desktop PC monitor (PC-LVO). To evaluate the accuracy of an arterial occlusion diagnosis, interdevice variability between Smartphone-LVO and PC-LVO was analyzed using κ statistics, and image interpretation time was compared between Smartphone-LVO and PC-LVO.
There was broad agreement between Smartphone-LVO and PC-LVO evaluations regarding the presence or absence of arterial occlusion (Reader 1: κ=0.94;
The results indicate the evaluation of neuroimages using a smartphone application can provide an accurate and timely diagnosis of anterior intracranial arterial occlusion that can be shared immediately with members of the stroke team to support the management of patients with hyperacute ischemic stroke.
In acute stroke management, early initiation is the most important factor in successful reperfusion therapy for both intravenous tissue plasminogen activator (IV-tPA) and mechanical thrombectomy (MT) [
Because “time is brain” in acute stroke treatment, the appropriate use of information and communication technology (ICT), particularly telemedicine for stroke or “telestroke,” has become a valuable method for reaching out to patients who are eligible for IV-tPA and MT administration. Telestroke is continually undergoing development, specifically in the United States and the European Union. The JOIN smartphone application enables the multidisciplinary stroke team to share clinical information and imaging data securely. Through JOIN, the intrahospital network system sends medical images from the hospital server to a shared chat room accessible to stroke team members only, similar to modified teleradiology and teleconferencing (
A. The JOIN smartphone application utilizes the easy-to-use interface of the social networking communication environment. B. The JOIN smartphone application displays diagnostic medical images, such as magnetic resonance imaging. ICA: internal carotid artery; MRA: magnetic resonance angiography.
Because the evaluation of large vessel occlusions (LVO) is vital for successful MT [
Consecutive patients admitted with ischemic stroke to Jikei University Hospital, Japan, between January 2016 and September 2017 were retrospectively enrolled in the study. The subtype of ischemic stroke was classified according to the TOAST (Trial of ORG 10172 in Acute Stroke Treatment) [
The JOIN (Allm, Tokyo, Japan) smartphone application enables teleradiology of medical images, including MRI, CT, ultrasonography, and angiography (
All patients underwent routine 3D time-of-flight (TOF) MRA of the intracranial arteries and fluid-attenuated inversion recovery (FLAIR) of the brain on a 1.5T scanner (MAGNETOM Avanto and MAGNETOM Symphony). The imaging parameters were TOF MRA: repetition time (TR)/ echo time (TE) = 23/7.15 ms, flip angle = 20°, section thickness = 0.6 mm, matrix = 320 × 288, field-of-view (FOV) = 18 cm; FLAIR: TR/TE = 9000/94 ms, flip angle = 160°, section thickness = 5 mm, section gap = 1.5 mm, matrix = 256 × 256, FOV = 21 cm.
The image review process is shown in
Flow diagram of the study design. LVO: large vessel occlusion, MRI: magnetic resonance imaging.
Interdevice agreement for evaluating the presence of arterial occlusion was analyzed between Smartphone-LVO and PC-LVO via kappa statistics. Kappa scores for agreement were rated as follows: poor (<0.20), fair (0.21-0.40), moderate (0.41-0.60), favorable (0.61-0.80), almost perfect (0.81-1.0).
Wilcoxon signed-rank test was used to compare image interpretation times between Smartphone-LVO and PC-LVO.
The Regional Ethics and Hospital Management Committee of the Jikei University School of Medicine approved the study (approval No. 29-197). The board waived the need for patient consent by giving patients from whom data had been collected the opportunity to opt-out of the study.
Patient characteristics.
Variable | All patients (n=108) | |
Age (years), median (IQR) | 69 (61-80) | |
Sex (male), n (%) | 72 (67) | |
NIHSSa score on admission, median (IQR) | 4 (2-7) | |
|
||
|
Large-artery atherosclerosis | 8 (7) |
|
Small-vessel occlusion | 14 (13) |
|
Cardio-embolism | 35 (32) |
|
Other determined etiology | 9 (8) |
|
Undetermined | 42 (39) |
Modified Rankin scale at 3 months, median (IQR) | 1 (1-3) | |
MRIc time from onset (minutes), median (IQR) | 244 (107-623) |
aNIHSS: National Institute Health Stroke Scale.
bTOAST: Trial of ORG 10172 in Acute Stroke Treatment.
cMRI: magnetic resonance imaging.
Interdevice agreement for evaluating the presence or absence of intracranial arterial occlusion was almost perfect between Smartphone-LVO and PC-LVO (Reader 1: κ=0.94, 95% CI 0.82-0.97;
Comparison of evaluation of the presence or absence of intracranial arterial occlusion between Smartphone-LVO and PC-LVO.
|
Smartphone-LVOa | PC-LVO | |
|
|
Occlusion | No occlusion |
|
|||
|
Occlusion | 33 | 2 |
|
No occlusion | 1 | 72 |
|
|||
|
Occlusion | 42 | 5 |
|
No occlusion | 1 | 60 |
aLVO: large vessel occlusion.
Interdevice agreement for evaluation of intracerebral arterial occlusion and for the 3 vessel categories. The degree of interdevice agreement was almost perfect in the evaluation of the arterial occlusion sites and was higher for proximal than distal arterial occlusion sites. ICA: internal carotid artery.
Comparison of identification of internal carotid artery occlusion between Smartphone-LVO and PC-LVO.
|
Smartphone-LVOa | PC-LVO | |
|
|
Occlusion | No occlusion |
|
|||
|
Occlusion | 8 | 0 |
|
No occlusion | 0 | 100 |
|
|||
|
Occlusion | 7 | 0 |
|
No occlusion | 0 | 101 |
aLVO: large vessel occlusion.
Comparison of identification of horizontal segment of middle cerebral artery occlusion between Smartphone-LVO and PC-LVO.
|
Smartphone-LVOa | PC-LVO | |
|
|
Occlusion | No occlusion |
|
|||
|
Occlusion | 16 | 1 |
|
No occlusion | 0 | 91 |
|
|||
|
Occlusion | 19 | 0 |
|
No occlusion | 4 | 85 |
aLVO: large vessel occlusion.
Comparison of identification of insular or cortical segments of middle cerebral artery occlusion between Smartphone-LVO and PC-LVO.
|
Smartphone-LVOa | PC-LVO | |
|
|
Occlusion | No occlusion |
|
|||
|
Occlusion | 31 | 4 |
|
No occlusion | 1 | 72 |
|
|||
|
Occlusion | 34 | 11 |
|
No occlusion | 2 | 61 |
aLVO: large vessel occlusion.
Wilcoxon signed-rank test (2-tailed) revealed that the interpretation time was significantly longer for Smartphone-LVO than PC-LVO (Reader 1: z=–3.547;
Comparison of interpretation times between Smartphone-LVO and PC-LVO.
|
Interpretation time of Smartphone-LVOa | Interpretation time of PC-LVO | ||
|
||||
|
Median minute (IQR) | 2(2-2) | 2(1-2) | |
|
Standard deviation | 0.64 | 0.68 | |
|
Min minute | 1 | 1 | |
|
Max minute | 4 | 3 | |
|
||||
|
Median minute (IQR) | 1(1-2) | 1(1-1) | |
|
Standard deviation | 0.49 | 0.34 | |
|
Min minute | 1 | 1 | |
|
Max minute | 3 | 2 |
aLVO: large vessel occlusion.
Difference in interpretation times between Smartphone-LVO and PC-LVO for Readers 1 and 2. The difference in interpretation time between Smartphone-LVO and PC-LVO was <1 min in 98% of all patient datasets. Reader 1, left; Reader 2, right. LVO: large vessel occlusion.
The study results show that the smartphone application yielded an accurate diagnosis of anterior intracranial arterial occlusion in times similar to diagnoses completed using the hospital-based desktop PC monitor. Several studies have reported the use of smartphone applications for the management of stroke [
The JOIN smartphone application shows promise for use in acute stroke management for several reasons. JOIN provides a platform for simultaneously sharing important clinical information among the entire stroke team via PACS, whether on duty or on call. Because the information shared within the team is secured, JOIN can seamlessly support intrahospital teleradiology for MT in the “drip and ship” acute stroke management model. Conventionally, after receiving a stroke call, the vascular neurologist must immediately access a desktop PC and log into the network and server before browsing images sent via PACS. The neurologist’s interpretation of the neuroimaging examinations is then shared over the telephone with one person in the stroke team. JOIN can substantially improve the speed of this process by reducing the time taken to evaluate neuroimaging, particularly in situations where the neurologist is not near a desktop PC or is not at the hospital at the time that treatment for hyperacute stroke is being decided. The utilization of the JOIN application should simplify and shorten the MT decision-making process.
Regarding the telestroke system, while peer-to-peer network protocols are standard, they are limited in that the sender communicates only with one receiver in the stroke team, much like a closed circuit loop. In contrast, the JOIN application enables multicasting communications from one sender to many receivers. The team leader pre-authorizes which stroke team members participate in the secure group chat, and all of the patient’s data are centralized to the group chat. The entire stroke team can then receive notifications of the stroke patient’s arrival, view neuroimaging examinations, and review the medical plan via JOIN. Thus, important information is shared immediately among all members. JOIN also enables messaging among the team and displays neuroimaging files at a high resolution. Because the decision-making process is tracked via messages displayed on JOIN, the application plays a role in avoiding errors. Thus, combining the conventional telestroke system with JOIN would be desirable for arranging hyperacute stroke care in a complex and time-sensitive environment.
Finally, the telestroke system, comprising the 3 components of teleconsultation (doctor to patient), teleconferencing, and teleradiology, is confronted with developmental challenges in Japan. It is difficult for hospitals in Japan to introduce the real-time interactive video conference systems required for tele-examination using a hub and spoke model because there is no legal or regulatory provision for doctors on the hub side regarding patient examinations or a provision for medical decisions [
Our study has several limitations. First, the MRA images were evaluated by only 2 vascular neurologists, both of whom were experienced. In a future study, more evaluators should be used, including less experienced evaluators. Second, the study did not directly prove that the smartphone application was useful for early reperfusion therapy. Finally, the study only evaluated anterior circulation ischemic stroke; its applicability to posterior circulation ischemic stroke has yet to be investigated.
A smartphone application yielded an accurate diagnosis of anterior intracranial arterial occlusion in times similar to diagnoses completed using the hospital-based desktop PC monitor. It also enabled access to all patient information regardless of the user’s location. Thus, this system can play an important role in organizing the stroke team to manage patients with hyperacute ischemic stroke.
The authors declare that all supporting data are available within the article.
fluid-attenuated inversion-recovery
field-of-view
hyperintense vessel sign
internal carotid artery
information and communication technology
intravenous tissue plasminogen activator
large vessel occlusion
magnetic resonance angiography
magnetic resonance imaging
mechanical thrombectomy
picture archiving and communication system
echo time
time-of-flight
repetition time
None declared.
TK contributed to the study design and concept, data acquisition and analysis, and drafted the manuscript for intellectual content. KS contributed substantially to the data acquisition. YI and YM revised the manuscript for intellectual content. HT and TI contributed to the data interpretation. YI approved the final version for publication.
HT and YM are inventors of the JOIN app. HT received private research funding from Allm, the developer of the JOIN app. The remaining authors declare no conflict of interest.