Decision support method for identification of the sciatic nerve in safe injection practices

Abstract

The invention relates to an artificial intelligence-based decision support method for determining the sciatic nerve prior to intramuscular injection and enabling safe injection.

Description

[0001] DECISION SUPPORT METHOD FOR IDENTIFICATION OF THE SCIATIC NERVE IN SAFE INJECTION PRACTICES

[0002] Technical Field

[0003] The invention relates to an artificial intelligence-based decision support method for determining the sciatic nerve prior to intramuscular injection and enabling safe injection.

[0004] Prior Art

[0005] Nurses are one of the important elements of the health systems of countries. According to the 2019 data of the Ministry of Health in our country; 198, 103 nurses are working. There are 430 patients per nurse in Turkey. (T.C. Saghk Bakanhgi, 2019

[0024] ). Parenteral drug administration is very important in medical treatment. Injection practice is among the most basic legally defined duties of nurses all over the world. The most commonly used injection site is the dorsogluteal region. However, its use is decreasing due to the increase in complications and the difficulty in determining the application area. Since there are sciatic nerve, superior gluteal nerve and artery in the dorsogluteal region, it is among the possible risks that injections applied to this region may damage these structures. Despite this, the dorsogluteal region is still recommended in some studies. (Kaya and ekin, 2018

[0011] ). Although intramuscular injections seem to be a very simple application, complications / errors cause the individual to experience significant health problems (Uysal & akircah, 2015

[0026] ). Although nurses receive adequate knowledge and skill training for drug administration and comply with the principles of drug administration, they may make drug administration errors. (Martyn & Paliadelis, 2019

[0013] ). In order to prevent drug administration errors, 10 correct principles (correct drug, correct dose, correct patient, correct time, correct route of administration, correct patient education, correct recording, confirmation of refusal, correct drug preparation and observation of whether the correct response develops in the patient) should be followed (Giindogdu, 2021[9]).

[0006] In intramuscular injection applications, the nurse's inability to determine the injection site correctly, incorrect dosing, use of non-sterile equipment, the nurse's lack of technical knowledge and skills, and some situations related to the patient may also cause errors (such as being old, weak) (Uslusoy et al., 2016

[0025] ). Medical errors related to nurses are mostly caused by situations such as excessive workload, carelessness, imprudence, inexperience in the profession, high number of patients per nurse and lack of communication (Er & Altunta§, 2016[6]). Medical mistakes made put nurses under legal responsibility in the face of the law (Uslusoy et al., 2016

[0025] ). Edema or scar tissue formed after injection, drug effect, individual anatomical differences, old age of the individual and cachectic are the predisposing factors for sciatic nerve damage (Kaya & ekin, 2018

[0011] ).

[0007] The most important complications seen in intramuscular injections are pain, abscess, necrosis, gluteal fibrosis, contracture, haematoma, periostitis and sciatic nerve injuries. As a result of sciatic nerve injury, neuropathy, lower extremity paralysis and foot deformities such as equinovarus occur (Geyik et al., 2017[7]; Alves et al., 2018a[1]; Alves et al., 2018b[2]; Riso et al., 2020

[0020] ; Mayer and Romain, 2001

[0014] ).

[0008] When the incidence and prevalence of sciatic nerve injury after gluteal intramuscular injection is examined; it is seen that it is a global health problem. (Mishra and Stringer, 2010

[0015] ; Park et al., 2019

[0017] ). In studies, it has been determined that sciatic nerve neuropathy develops at a rate of 31.2% in Turkey (Uslusoy et al., 2016

[0025] ) and 89.7% in India after intramuscular injection (Tak et al., 2008

[0023] ). In Pakistan, the estimated annual incidence of traumatic injection neuropathy is 7.1 per 1 million children under three years of age. (Mansoor et al., 2005

[0012] ). In studies conducted in our country, it has been determined that more than half of the patients with sciatic nerve injury have significant permanent damage (Pazarci et al., 2010

[0018] ; Geyik et al., 2017[7]).

[0009] The sciatic nerve is the largest nerve in the human body and is formed by the union of 5 nerve roots coming from the lower spine. It passes deep into the buttocks and runs from the back of the thigh to the heel and sole of the foot. The sciatic nerve plays a vital role in connecting the spinal cord to the skin and muscles of the thigh, leg and foot (Bagi§ et al., 2012[3]).

[0010] It is not possible to feel and see the sciatic nerve from the outside with sensory organs (such as hand, eye). Therefore, the introduction of ultrasound has led to very important developments in medicine. The ability to see the nerves, surrounding tissues and local anaesthetic distribution with ultrasound has increased the quality and success rate of sciatic nerve block and reduced the frequency and severity of complications. Some medical applications that are difficult to perform due to anatomical variations have been facilitated by ultrasound (§ahin et al., 2011

[0021] ).

[0011] In the literature, it is stated that sciatic nerve movements can be seen and imaged with different positions given to the individual with ultrasound. In a study, it was emphasised that real-time ultrasound imaging may be useful in sciatic nerve blocks to facilitate sciatic nerve localisation (Balaban et al., 2020[4]). In another study, it was suggested that ultrasound-guided peripheral nerve blocks can be applied effectively and safely in early postoperative pain in orthopedic paediatric surgery and regional blocks in paediatric patients (§ahin et al., 2011

[0022] ).

[0012] When other studies are examined; it is understood that ultrasound is used in sciatic nerve imaging and sciatic nerve block (Karmakar et al., 2012

[0010] ; Mori and Hagiwara, 2017

[0016] ; Goldsmith et al., 2020[8]).

[0013] Technology has been developing rapidly in the world recently. As in other sectors, the health sector is affected by these developments. In our country, there is a transition to decision support systems in hospitals in terms of adaptation and innovation to technological developments in the field of health. (Purkuluoglu et al., 2019

[0019] ). In recent years, nurses have been using decision support systems to provide fast, safe, effective, evidence-based and quality sustainable care to their patients. It is seen that decision support systems are mostly used in laboratory requests, prescribing medication, adjusting drug doses and giving instructions. Decision support systems in hospitals reduce costs and reduce diagnosis, treatment and care errors. In the future, clinical decision support systems will provide high quality nursing services with low cost and few errors (Purkuluoglu et al., 2019

[0019] ).

[0014] Intramuscular injection has many risks. The most commonly used intramuscular injection site by nurses in our country is the dorsa gluteal region. It is stated that this region is the most risky region for intramuscular injection because it is rich in vessels, close to the sciatic nerve and the subcutaneous tissue is thicker than the others. In the study conducted with 110 nurses in the technique, it was found that 60% of the nurses always used the dorsa gluteal region in injection applications, while 78.2% never used the ventrogluteal region. Sciatic nerve is the longest and thickest nerve of the human body. Sciatic nerve variations vary from individual to individual. Injection directly into the nerve, pressure of the drugs into the area and chemical effects of the drugs cause sciatic nerve injury. In the first method, the posterior superior iliac spina and the greater trochanter of the femur are joined by an imaginary line. A suitable area is selected above this line and below the iliac crista. In the second method, the right / left hip is divided into 4 equal parts by horizontal and vertical lines. The upper outer part is divided into four again. A suitable area is selected in the top and outer region. In the third method, the crista iliacus superior and the coccyx are connected with an imaginary line and divided into three. The appropriate area is selected at the outer 1 / 3 point. (Karabacak BG

[0027] ) Injection continues to be performed with the methods described above.

[0015] Complications such as abscess, necrosis, infection, tissue irritation, contracture, haematoma, chronic pain, periostitis, injury to vessels, bones and nerves may develop due to intramuscular injections performed by nurses with the technique they know. Sciatic nerve injury may occur especially due to injections in the dorsagluteal region. As a result of sciatic nerve injury, foot drop, loss of flexion and extension of the toes, loss of sensation, pain in the legs and feet and loss of sensitivity in the feet may develop. (Kaya and ekin, 2018

[0011] ).

[0016] CN210228608U discloses a stool for accurately positioning the injection position.

[0017] 2016 / 04323 discloses a detection apparatus developed to determine the injection site where intramuscular drug administration is performed on the ventrogluteal region and where the muscle is the densest.

[0018] When the studies in the prior art are examined, it has been necessary to develop an artificial intelligence-based decision support method that enables safe injection by determining the sciatic nerve before the intramuscular injection subject to the invention in order to solve the problems existing in the technique.

[0019] Objectives of the Invention

[0020] The object of the present invention is to develop an artificial intelligence based decision support method that enables safe injection by identifying the sciatic nerve before intramuscular injection.

[0021] Another object of the present invention is to develop an artificial intelligence-based decision support method that will reduce the risk of medical errors in injection practices of healthcare personnel such as doctors, nurses and midwives.

[0022] Detailed Description of the Invention.

[0023] This patent application relates to an artificial intelligence-based decision support method for the safe identification of the sciatic nerve prior to intramuscular injection. This method includes the following steps:

[0024] 1. Taking images of the sciatic nerve using an ultrasonic probe. 2. The images are converted to JPEG format for processing.

[0025] 3. Detecting of sciatic nerves with scaled deep learning based single frame object detection method by preprocessing the data in JPEG format.

[0026] This method, ultrasonography-guided imaging of the sciatic nerve, was developed using a variety of patient data in different positions and situations. The images are obtained using ultrasound technology that can easily follow the course of the nerve, and the deep learning model is integrated and notified to the experts.

[0027] In particular, supervised learning, transfer learning and the use of pre-trained models play an important role in the training of the deep learning model. The data preprocessing stage is an important step for cleaning and harmonising the raw data.

[0028] As a result, this patent application describes an innovative decision support method that combines artificial intelligence and ultrasound technologies to safely identify the sciatic nerve prior to intramuscular injection.

[0029] This methodology uses a convex probe (5-7.5 MHz) under ultrasound guidance as the first step in the imaging of the sciatic nerve. B-mode (grey scale) images of the sciatic nerve will be obtained in axial and sagittal planes in lateral decubitus, supine, and prone positions with the thigh in neutral position and flexion on a total of 200 patients of different ages and genders. These images will be taken in sagittal and axial sections to include the entire course of the nerve in the gluteal region. The images will be converted to jpeg format and saved in a database.

[0030] Ultrasound imaging technology has made the needles and target nerve structures visible. Nerve recognition can be confirmed by scanning the nerve along its known course. Ultrasound can follow the course of nerves more easily than other imaging methods. At this stage, the first step for integrating the deep learning model is to compute the sciatic nerve with deep learning algorithms and notify the expert. The aim is to provide a radiology-based interpretation of the ultrasound images, in particular the automatic detection of the location of the sciatic nerve. Images taken at different times will be collected to develop a deep learning based decision support algorithm using real time data. These data will be checked by radiology specialists and appropriate measures will be taken to solve the problem of unstable data. In addition, weighted loss calculations or data augmentation methods will be used in case of imbalanced health data.

[0031] In the data preprocessing stage, the received data will be analysed and the blurred and structurally inappropriate images will be cleaned, and the unprocessed data stacks caused by the data being received in different situations will be examined and made appropriate. Data preprocessing is a critical step that helps to build machine learning models more accurately. In the training phase, it is planned to train the health dataset with pre-trained models and apply transfer learning.

[0032] In model training, different learning algorithms, especially F-CNN, R-CNN, Faster R-CNN, Faster R-CNN, Y0L0v4, Y0L0v5, and Scaled- Y0L0v4, and a model developed and adapted to ultrasonic systems will be evaluated on training data, validation data and test data and their performances will be examined.

[0033] References:

[0034] 1. Alves, K., Godwin, C. L., Chen, A., Akellot, D., Katz, J. N., Sabatini, C. S. 2018a. “Gluteal fibrosis, post-injection paralysis, and related injection practices in Uganda: a qualitative analysis”, BMC Health Services Research, 18:892,1-11.

[0035] 2. Alves, K., Penny, N., Ekure, J., Olupot, R., Kobusingye, O., Katz, J. N., Sabatini, C. S. 2018b. “Burden of gluteal fibrosis and post-injection paralysis in the children of Kumi District in Uganda”, BMC Musculoskelet Disorder, 19:343,1-6.

[0036] 3. Bagi§, S., Adam, M., Leblebici, U. B., Karata§, M., Guven, A. Z., eliker, A. R. 2012. “Sciatic nerve injury due to intramuscular injection: electrophysiological findings and one-year follow-up”, Turkish Journal of Medical Sciences, 42(5), 913-17. Balaban, O., Yaman, M., Aydin, T., Musmul, A. 2020. “Ultrasound detection of sciatic nerve movements with ankle dorsiflexion / plantar flexion: prospective comparative study of a novel method to locate the sciatic nerve”, The Journal of the Turkish Society of Algology, 32(3), 152- 8. Bilgici, A., okluk, C., Aydin, K. 2013. “Ultrasound neurography in the evaluation of sciatic nerve injuries”, Journal of Physical Therapy Science, 25(10), 1209-11. Er, F., Altunta§, S. 2016. “Hem§irelerin tibbi hata yapma durumlan ve nedenlerine ybnelik goru§lerinin belirlenmesi”, Saghk ve Hem§irelik Ybnetimi Dergisi, 3(3), 132-9. Geyik, S., Geyik, M., Yigiter, R., Kuzudi§li, S., Saglam, S., Elgi, MA., Yilmaz, M. 2017. “Preventing sciatic nerve injury due to intramuscular injection: ten-year single-center experience and literature review”, Turk Neurosurgery, 27(4), 636-40. Goldsmith, A. J., Liteplo, A., Hayes, B. D., Duggan, N., Huang, C., Shokoohi, H. 2020. “Ultrasound-guided transgluteal sciatic nerve analgesia for refractory back pain in the ED”, The American Journal of Emergency Medicine, 38(9), 1792-95. Giindogdu, H. 2021. Hem§irelik Temel Becerilerinde Klinik Notlar "cep kitabi", Ed. Yilmaz D, Dikmen Y.1. Baski, Vize Yayincihk, Ankara, ss.39- 57. Karmakar, M., Li, X., Li, J., Sala-Blanch, X., Hadzic, A., & Gin, T. 2012. “Three-dimensional / four-dimensional volumetric ultrasound imaging of the sciatic nerve”, Regional Anesthesia and Pain Medicine, 37(1), 60-6. Kaya, K., ekin, N. 2018. “Enjeksiyon sonrasi geli§en noropati: komplikasyon / malpraktis aynminda ince bir gizgi”, KSU Medical Journal, 13(2), 63-6. 12. Mansoor, F., Hamid, S., Mir, T., Abdul Hafiz, R., Mounts, A. 2005. “Incidence of traumatic injection neuropathy among children in Pakistan”, EMHJ-Eastern Mediterranean Health Journal, 11(4), 798-804.

[0037] 13. Martyn, J. A., Paliadelis, P. 2019. “Safe medication administration: Perspectives from an appreciative inquiry of the practice of registered nurses in regional Australia”, Nurse Education in Practice, 34, 111-6.

[0038] 14. Mayer, M., Romain, O. 2001. “Sciatic paralysis after a buttock intramuscular injection in children: an ongoing risk factor”, Archives De Pediatrie: Organe Officiel De La Societe Francaise De Pediatrie, 8(3), 321- 3.

[0039] 15. Mishra, P., Stringer, M. D. 2010. “Sciatic nerve injury from intramuscular injection: a persistent and global problem”, International Journal of Clinical Practice, 64(11), 1573-9.

[0040] 16. Mori, T., Hagiwara, Y. 2017. “Ultrasound-Guided popliteal sciatic nerve block for an ankle laceration in a pediatric emergency department”, Pediatric Emergency Care, 33(12), 803-5.

[0041] 17. Park, C. W., Cho, W. C., Son, B. C. 2019. “Iatrogenic injury to the sciatic nerve due to intramuscular injection: a case report”, Korean Journal of Neurotrauma, 15(1), 61-6.

[0042] 18. Pazarci, N. K., Orken, D. N., elik, M. G., elebi, L. G., Aydin, S. 2010. “Postenjeksiyon siyatik noropati: klinik ve elektrofizyolojik ozellikler”, Noro-Psikyatri Ar§ivi, 47(3), 207-12.

[0043] 19. Purkuloglu, E., Un, A., Yuriir Durmaz, F. 2019. “Hem§ire karar destek sistemleri uygulamalan”, Hacettepe Saghk idaresi Dergisi, 22(3), 491-514.

[0044] 20. Riso, V., lodice, F., Barbato, F., Granata, G. 2020. “Ultrasound aspects of neurotoxicity-mediated sciatic nerve injury after intramuscular ketorolac injection”, Journal of Ultrasound in Medicine: Official Journal of The American Institute of Ultrasound in Medicine, 39(9), 1877-8. 21. §ahin, L., §ahin, M., Akta§, O., Gill, R. 2011. “Ortopedik cerrahi igin pediatrik hastalarda ultrason klavuzlugunda rejyonal anestezi”, Gaziantep Medical Journal, 17(2), 63-6.

[0045] 22. §ahin, L., §ahin, M., Gill, R. 2011. “Ortopedik cerrahi sonrasi pediatrik olguda bilateral siyatik sinir blogu”, Dicle Tip Dergisi, 38(3), 355-57.

[0046] 23. Tak, S. R., Dar, G. N., Halwai, M. A., Mir, M. R. 2008. “Post-injection nerve injuries in Kashmir: A menace overlooked”, Journal of Research in Medical Sciences, 13(5), 244-7.

[0047] 24. Tiirkiye Cumhuriyeti Saghk Bakanhgi. “Saghk Istatistikleri Yilhgi 2018”. https: / / dosyasb. saglik.gov. tr / Eklenti / 36134,siy2018trpdf.pdf?0. Son eri§im tarihi: 12 Arahk 2020.

[0048] 25. Uslusoy, E. ., Duran, E. T., Korkmaz, M. 2016. “Giivenli enjeksiyon uygulamalan”, Hacettepe Universitesi Hem§irelik Fakiiltesi Dergisi, 3(2), 50-7. 26. Uysal, N., akircah, E. 2015. Hem§irelik Esaslan insan Saghgi ve

[0049] Fonksiyonlan (7.Basim). Ankara: Palme Yayincihk.

[0050] 27. Karabacak BG. Parenteral ilag Uygulamalan iginde Klinik Beceriler: Saghgin Degerlendirilmesi, Hasta Bakim ve Takibi. Eds: Sabuncu N, Ay FA. Nobel Tip Kitabevleri, Istanbul 2010, 250-300.

Claims

CLAIMS1. An artificial intelligence based decision support method for determining the sciatic nerve prior to intramuscular injection and enabling safe injection, charactrized in that it comprises, - Taking sciatic nerve images with the help of ultrasonic probe,Converting the received images to jpeg format for processing,- Pre-processing of the data obtained in Jpeg format,- Detecting of sciatic nerves by processing the pre-processed and cleaned data with the scaled-YOLOv4 method.

2. An artificial intelligence based decision support method according to claim1, characterized in that deep learning -based single frame object detection methods can be used instead of the scaled Y0L0v4 method.