Uploaded by givsonmc

[White paper] NerveTrack™ WP202110

WHITE PAPER
Automatic Nerve Tracing Ultrasound
Technology in Pain Practice
Jee Youn Moon, MD, PhD, FIPP, CIPS.
Department of Anesthesiology and Pain Medicine,
Seoul National University Hospital College of Medicine.
101 Daehak-ro, Jongno-gu, Seoul 03080, Republic of Korea
Introduction
According to the International Association of Pain, Pain is defined as “an unpleasant sensory
and emotional experience associated with, or resembling that associated with, actual or
potential tissue damage” [1]. Therefore, pain is thought as a unique and personalized symptom [1, 2]. Regarding pain evaluation, pain physicians frequently see patients who complain
of severe low back pain without any significant lesion corresponding to their pain severity in
MRI or other imaging workups. These cases might be so-called somatoform [2] or even
malingering in the past; however, do pain physicians verify that it is not real pain? Further,
even if pain physicians eliminated pain causes in the prognosis, the symptom could not be
totally relieved (chronification), resulting in secondary catastrophic adverse events, such as
depressive mood and insomnia [2, 3]. Due to these subjective and unique pain characteristics, it is always challenging to evaluate and manage pain. Therefore, it is a mission for pain
physicians to begin pain management with acknowledgment and empathy to their patients.
Despite the pain subjectivity, modern technology has remarkably contributed to shaping
objective ways to diagnose and manage various pain disorders. Best of all, Artificial Intelligence (AI) has been pivotal in it [4]. AI technology and machine learning have suggested
algorithms to predict a pain phenotype class from a complex pattern of acquired parameters
[4] and predictors of response to pregabalin for broad neuropathic pain [5]. Further, the
effectiveness of AI in medical image interpretation is well-documented [6]. Pain physicians
can benefit from AI technology during image-guided pain procedures. Ultrasonography (US)
is one of the most efficient devices in adopting AI technology. In pain practice, US-guided
joint injections, interfascial blocks, peripheral nerve blocks, and even neuraxial blocks have
been popular since the early 2000s [7, 8]. Compared to fluoroscopy, US has several practical
advantages, such as no radiation exposure, possible dynamic evaluation, and easy portability
[8]. Therefore, it has been essential for pain fellows and trainees to study sonoanatomy to
perform US-guided pain procedures. Peripheral nerve blocks under US guidance are commonly conducted for regional analgesia and pain management. For a successful procedure,
start scanning the target nerve by tracing it with the probe, followed next by scanning the
nerve pathway. This allows clinicians to examine the target nerve path accompanied by the
adjacent muscles, ligaments, tendons, vessels, and bone. Without the pre-scanning step,
clinicians may face ambiguous cases when distinguishing between nerves and tendons, as
encountered with injection of the median nerve block in the carpal tunnel. Furthermore, the
pre-scanning exam may be challenging for trainees and beginners to recognize the target
nerve without an expert’s supervision.
2
What if the AI technology could assist us to trace those target nerves during the US scanning
and even point to the target instantaneously when the probe is placed near the target structure? The NerveTrackTM developed by SAMSUNG MEDISON is a cutting-edge technology to
assist clinicians during pain procedures and to intelligently find the target structure, especially nerves such as the median nerve, the ulnar nerve, etc. It is a built-in, commercially
available software installed on Samsung’s high-resolution US systems (SAMSUNG MEDISON,
CO., LTD. Republic of Korea). It was created using AI deep learning technology to read hundreds of thousands of refined US images for each targeted nerve. By simply placing the US
probe near the target area, NerveTrackTM can find the target nerve, automatically skipping
the pre-scanning step. Further, it can point out the nerve instantaneously by tracing it. NerveTrackTM is expected to help clinicians and trainees perform effective and correct procedures under US guidance, reducing the procedure time and enhancing the patients’ safety by
avoiding injuries to the adjacent vessels and tendons. Additionally, this technology may contribute to education in pain medicine by assisting trainees, residents, and pain fellows to
learn to recognize the target structure, reducing mistakes in practice as they are making
medical decisions during procedures.
Here, we introduce a virtual case detecting the median nerve using the NerveTrackTM technology along with interim results of a survey regarding the technology answered by residents in the Department of Anesthesiology and Pain Medicine and 2nd and 3rd-year medical
students at Seoul National University Hospital College of Medicine.
3
Case and Survey
The median nerve can be entrapped in several locations in the upper extremity. They include
areas between the humeral and ulnar heads of the pronator teres muscle, under the fibrous
arcade of the flexor digitorum superficialis, at the lacertus fibrosus of biceps aponeurosis or
the Struthers’ ligament, and most commonly under the transverse carpal ligament –also
known as carpal tunnel syndrome (CTS) [9]. CTS is a prevalent entrapment disorder; the
lifetime prevalence is 7.8%, higher for women (10.0%) than men (5.8%), depending on occupational risk [10]. Diagnosis of CTS is based on signs and symptoms found during a clinical
exam with or without electrophysiologic studies [9-11]. Additionally, US has been frequently
used to examine the morphology of the median nerve and guide the needle to the median
nerve [10, 11]. During the US examination, enlargement of the median nerve seems to be the
most sensitive and specific criterion, accompanied by a palmar bowing of the flexor retinaculum and distal flattening of the nerve [11].
The virtual patient was seated, with the forearm supinated and the wrist placed in slight
dorsiflexion using a rolled towel. As the first step of the US examination, a transverse sonogram of the carpal tunnel was obtained by placing a high-frequency linear transducer on the
proximal wrist crease at the entrance to the carpal tunnel. Within the carpal tunnel, we identified nine tendinous structures (four flexor digitorum superficialis, four flexor digitorum
profundus, and flexor pollicis longus tendons), which had boundaries from the pisiform (medially) and scaphoid (laterally) [11]. They were covered by the transverse carpal ligament
(Figure 1).
Medial
Figure 1. Transverse ultrasound image of
the proximal carpal tunnel.
The median nerve is shown at the middle
top of the image with a yellow arrow.
4
Although identifying the median nerve under the proximal carpal tunnel is an easy technique for experts, it may confuse beginners. In 40 residents and medical students who were
not familiar with the use of US in pain practice, we surveyed whether these beginners could
point out the median nerve after an observational US tracing step without additional education. Only 6 participants pointed the nerve out correctly; however, they were still confused,
did not display confidence and could not explain the other structures around the target. The
remaining 34 trainees confused the median nerve with either the flexor pollicis longus
tendon (N = 15), the flexor carpi radialis tendon (N = 13), one of the flexor digitorum profundus tendons (N = 4), one of the flexor digitorum superficialis tendons (N = 1), and the flexor
carpi ulnaris tendon (N = 1). Next, we educated them, showing the median nerve traced at the
forearm, and performed the same test. Nineteen participants correctly identified the nerve
at the proximal carpal tunnel. Then, we educated them on activating the NerveTrackTM function (Figure 2A, 2B, 2C, and 2D).
At this point, 37 of the trainees pointed out the median nerve correctly at any location in the
forearm without confusion or hesitancy. All participants indicated that the NerveTrackTM
helped extremely (N = 29) or helped very much (N = 11) in their education. Before the education using the NerveTrackTM, their certainty score for correctly identifying the median nerve,
which was described from 1 (not confident at all) to 5 (completely confident), ranged from 1
to 2 (slightly confident). After the education using the NerveTrackTM, their confidence to
detect the median nerve even without activation of NerveTrackTM soared to 3 (N = 7) somewhat confident, 4 (N = 26) fairly confident, and 5 (N = 5) completely confident. Two participants responded that they were not confident using NerveTrackTM.
5
Figure 2. Transverse ultrasound images tracing the median nerve using the NerveTrackTM
Medial
Lateral
Medial
Lateral
FCR
FDS
FDS
FDP
PQ
FDP
Figure 2A. Median nerve at the forearm. The nerve is identi-
Figure 2B. The median nerve proximal at the forearm before
fied between the flexor digitorum superficialis (FDS) and
it entered the proximal carpal tunnel. The nerve is identified
flexor digitorum profundus (FDP) muscles. The median
between the flexor digitorum superficialis (FDS) and flexor
nerve is identified within a yellow box.
digitorum profundus (FDP) muscles. Additionally, flexor
carpi radialis (FCR) and pronator quadratus (PQ) are
adjacently identified in the image. The median nerve is
identified within a yellow box.
Medial
FCR
FDS
FDP
FCR
FDS
FPL
Lateral
FPL
FDF
Figure 2C. The median nerve at the proximal carpal tunnel.
Figure 2D. The median nerve at the distal carpal tunnel. The
The median nerve is identified within a yellow box beneath
median nerve in the yellow box runs deeply into the carpal
the transverse carpal ligament. There are flexor digitorum
tunnel.
superficialis (FDS) tendons, flexor digitorum profundus
(FDP) tendons, and flexor pollicis longus (FPL) tendon within
the carpal tunnel. The flexor carpi radialis (FCR) tendon is
shown outside of the tunnel. The median nerve is identified
within a yellow box.
6
Medial
Closing
Below are suggested impacts of AI on pain medicine. The AI and deep learning technology
have assisted in:
1. Pain diagnosis and its evaluation, e.g., suggesting the ability to detect patterns of clinical characteristics in various phenotypes of chronic pain;
2. Pain management, e.g., when physicians perform ultrasound-guided procedures, AI
technology assists pain physicians to recognize the target structures more efficiently to
avoid damages to other critical structures, such as when using the NerveTrackTM technology; and
3. Education in pain medicine, e.g., helping trainees, residents, and pain fellows to learn the
process of making diagnoses and finding the targeted structure during their procedures,
as described in our survey results.
This AI technology will improve clinical performance and decrease mistakes in pain practice.
Some theorize that doctors will be replaced by machines using AI technology in the future.
However, this appears to be an obfuscation so far and in fact, the opposite is occurring presently. The availability of big data has already driven the use of AI in healthcare to train predictive algorithms, which helps human doctors (instead of replacing them), encourages curiosity-based reasoning, empowers collaboration and eliminates unremarkable tasks. We
know that medicine must be developed with technological advancement as well as humanism. Artificial Intelligence in healthcare will improve this technical advancement and
humanism in clinical diagnosis and decision-making by automating patient assessments and
eliminating assessor bias.
Supported system - HS40
NerveTrackTM is available in the following systems - V8, HS40
7
References
1. Raja SN, Carr DB, Cohen M, et al. The revised International Association for the Study
of Pain definition of pain: concepts challenges, and compromises. Pain.
2020;161:1976-82.
2. Treed RD, Rief W, Berke A, et al. Chronic pain as a symptom or a disease: the IASP classification of chronic pain for the International Classification of Diseases (ICD-11).
Pain. 2019;160:19-27.
3. Malfliet A, Coppieters I, Van Wilgen P, et al. Brain changes associated with cognitive
and emotional factors in chronic pain: A systematic review. Eur J Pain.
2017;21:769-86.
4. Lotsch J, Ultsch A. Machine learning in pain research. Pain.
2018;159:623-30.
5. Emir B, Kuhn M, Johnson K, et al. Predictors of response to pregabalin for broad neuropathic pain: results from 11 machine leaning methods from a 6-week German
observational study. J Pain. 2016;17:S78.
6. Hosny A, Parmar C, Quackenbush J, et al. Artificial intelligence in radiology. Nat Rev
Cancer. 2018;18:500-10.
7. Krediet AC, Moayeri N, van Geffen GJ, et al. Different approaches to ultrasound-guided thoracic paravertebral block: An illustrated review. Anesthesiology.
2015;123:459-74.
8. Ryu JH, Lee CS, Kim YC, et al. Ultrasound-assisted versus fluoroscopic-guided lumbar
sympathetic ganglion block: A prospective and randomized study. Anesth Analg.
2018;126:1362-1368.
9. Doughty CT, Bowley MP. Entrapment neuropathies of the upper extremity. Med Clin
North Am. 2019;103:357-70.
10. Prevalence and incidence of carpal tunnel syndrome in US working populations:
pooled analysis of six prospective studies. Scand J Work Environ Health,
2013;39:495-505.
11. Erickson M, Lawrence M, Lucado A. The role of diagnostic ultrasound in the examination of carpal tunnel syndrome: an update and systematic review. J Hand Ther.
2021;14:S0894-1130(21)00061-2. Online ahead of print.
8
Disclaimer
* The features mentioned in this document may not be commercially available in all countries. Due to regulatory reasons, their future
availability cannot be guaranteed. Please contact your local sales network for further details.
* Do not distribute this document to customers unless relevant regulatory and legal affairs officers approve such distribution.
* Images may have been cropped to better visualize its pathology.
* This clinical practice review is not an official clinical study or paper presented at a conference. It is a result of a personal study conducted by
collaboration between Samsung Medison and Prof. Jee Youn Moon. This review is to aid customer in their understanding, but the objectivity
SAMSUNG MEDISON CO., LTD.
© 2021 Samsung Medison All Rights Reserved.
Samsung Medison reserves the right to modify any design, packaging,
https://www.samsunghealthcare.com/en/products/UltrasoundSystem
WP202110-NerveTrack™
is not secured.