Ultrasound Radiomics-Guided Iliac Fascia Block on Postoperative Cognitive Dysfunction in Elderly Patients Undergoing Hip Surgery

Zha, Jun; Ni, Jinping; Chen, Shuo; Feng, Haijun; Che, Tuanjie; Qiao, Shigang

doi:https://doi.org/10.1155/2022/2492667

Computational and Mathematical Methods in Medicine

On this page

Abstract Introduction Materials and Methods Results Discussion Conclusion Data Availability Conflicts of Interest Authors’ Contributions Acknowledgments References Copyright Related Articles

Special Issue

Machine Learning for Human-Machine Interaction

View this Special Issue

Research Article | Open Access

Volume 2022 | Article ID 2492667 | https://doi.org/10.1155/2022/2492667

Ultrasound Radiomics-Guided Iliac Fascia Block on Postoperative Cognitive Dysfunction in Elderly Patients Undergoing Hip Surgery

Jun Zha,¹Jinping Ni,²Shuo Chen,¹Haijun Feng,¹Tuanjie Che,^3,4and Shigang Qiao¹

Academic Editor: Kelvin Wong

Received20 Jan 2022

Revised25 Mar 2022

Accepted26 Apr 2022

Published17 May 2022

Abstract

Objective. Elderly patients with hip surgery are prone to postoperative cognitive dysfunction (POCD), leading to health management difficulties. This study is aimed at investigating the effect of ultrasound radiomics-guided iliac fascia block on POCD. Methods. A total of 67 cases of patients who had undergone hip joint surgery were divided into a training set () and a validation set (radiomics-guided group, ). The patients were intervened with ultrasound radiomics-guided iliac fascia block, and the maximum relevance minimum redundancy sifts out the image omics features obtained from 2D ultrasound images of patients. Another 20 patients undergone general anesthesia served as control. The incidence of POCD, the total amount of fentanyl, the visual analogue score (VAS) at different time points, and the levels of CRP and NSE in plasma were compared between the two groups. Results. The AUC on the training and validation sets were higher than 0.940. The incidence of POCD in the radiomics-guided and general anesthesia group was 5% and 30%, respectively (). Compared with the general anesthesia group, the dosage of fentanyl in the radiomics-guided was lower, the VAS score at 6 h, 1 d, and 2 d after operation was smaller, and the levels of CRP and NSE were lower (all ). Conclusions. For elderly patients with hip surgery, the ultrasound radiomics-guided iliac fascia block can reduce the incidence of POCD and improve the effect of nerve block.

1. Introduction

The postoperative cognitive dysfunction (POCD) often occurs in elderly patients; it refers to a series of neurological complication after surgery [1–3]. Early intervention can reduce the incidence of POCD and minimize the degree of brain injury [4, 5]. Therefore, intervention in the influencing factors of POCD is an effective measure to reduce the occurrence of POCD.

In addition to old age, pain is also considered a high risk factor for POCD [6–9]. The mechanism of central inflammatory response also plays an essential role [10]. Patients with hip disease not only have severe pain before the operation but also suffer from pain interference after the operation [11–14], resulting in abnormal difficulty in health management. Therefore, appropriate preoperative and postoperative analgesia may be helpful for the control of POCD.

In the surgical treatment of hip joint lesions, some studies have shown that iliac fascia block can play a better analgesic effect and reduce postoperative complications [15–19]. However, it is not clear whether POCD can be affected. The radiomics features are of great value in diagnosing clinical diseases and evaluating treatment response [20–22]. In recent years, the clinical application has gradually increased. In this study, based on iliac fascia block ultrasound radiomics-guided, we tried to reduce the impact of surgery on COPD in addition to reducing pain.

2. Materials and Methods

2.1. General Information

All the participants were patients in Suzhou Science and Technology Town Hospital and Suzhou Kowloon Hospital for elective surgery in patients with hip disease. The research design was approved by the ethics committee of our hospital, and the operation plan was informed consent of patients and their families.

Inclusive criteria are as follows: hip joint surgery patients, normal nerve function can cooperate with treatment, and years. Exclusion criteria are as follows: patients with a history of mental or neurological diseases, patients with abnormal visual or auditory function, patients with severe organ abnormalities, and patients with missing clinical and follow-up data or voluntarily applying to withdraw from the study.

Finally, 87 patients were enrolled in the study, including 55 males and 32 females. Among then, 67 cases were divided into a training set () and a validation set (radiomics-guided group, ) according to the 10-fold crossover method. Another 20 patients undergone general anesthesia served as the control group.

2.2. Anesthesia Method

The general anesthesia group was given sufentanil 0.5 μg/kg, propofol medium/long chain fat emulsion 2.5 mg/kg, rocuronium 0.8 mg/kg intravenous anesthesia induction, after the human pump propofol 6 mg/kg/h, and 0.8 MAC sevoflurane to maintain anesthesia.

The radiomics-guided group was blocked in the iliac fascia space: supine, the affected side of the limb slightly extended outward, conventional disinfection spread sterile towel, the Sonosite S-nerve color Doppler was used. After the probe was placed vertically at the position of anterior superior iliac spine, it was rotated about 20° clockwise and then moved slowly to the middle and outer 1/3 of the anterior superior iliac spine and pubic symphysis. The probe was adjusted to determine the structure of the iliac fascia space. When the needle tip penetrates below the iliac fascia, inject 30 mL of 0.25% ropivacaine hydrochloride after no blood is drawn back, pay attention to the diffusion of local anesthetic to the head, and keep the needle route within the visual range to prevent the needle from entering the abdominal cavity too deep.

2.3. Postoperative Analgesia

Patients in both sets were given patient-controlled intravenous analgesia. Fentanyl (20 μg/kg) was dissolved in 100 mL 0.9% sodium chloride solution. The injection speed was controlled at 2 mL/h and locked for 15 min. The single self-control volume was 0.5 mL.

2.4. Image Preprocessing and Region of Interest Segmentation

Image preprocessing can be divided into two steps: image gray balance and image noise reduction. Histogram equalization reduces feature differences caused by image gain differences between patients. To reduce the influence of image noise on image omics feature extraction, median filtering is applied to image denoising.

Image segmentation was performed manually by an ultrasound physician. Region of interest (ROI) was plotted on the maximum cross-section of the injected ultrasound image of each patient. The complete structural range of iliac fascia was plotted by adjusting the spatial structure position. The segmented images were finally validated by an ultrasound physician with 5 years of experience. The specific flow chart for image preprocessing and region of interest segmentation is shown in Figure 1.

2.5. Radiomics Analysis

The deep learning model in this research is generated by the Deep Red AI Toolbox. Firstly, the image omics features of the region of interest are extracted from the preprocessed images. Then, the extracted features are normalized to make the data normalized and conform to the standard normal distribution. The transformation function is as follows:

In addition, the maximum relevance minimum redundancy (mRMR) and the least absolute shrinkage and selection operation (LASSO) algorithm sift out the image omics features obtained from 2D ultrasound images of patients and establish the image omics score.

The mRMR algorithm is a feature selection method that maximizes the correlation between feature variables and targets and minimizes the correlation between features. More mutual information is used to measure the feature redundancy and the correlation between categories, and information entropy and information difference are used to select the optimal feature subset.

In the formula, and are given two random variables, is the joint probability distribution function of and , and and are probability distributions of and , respectively.

To find the feature subset containing features, the following formula can be used to calculate its correlation:

A subset with features may not be the best subset of features, and when two features are highly interdependent and one of them is removed, the class discrimination of the two will not change much.

The selection of variables is mainly based on the LASSO algorithm proposed by Tibshirani et al. in 1996 [23]. The loss function is as follows: where is the category label of the th sample and is the number of samples. λ is the regularization parameter, which controls the sparsity of the model.

2.6. Evaluation Index

The incidence of POCD, the operation time, and the total amount of fentanyl were recorded.

The visual analogue score (VAS) was used to evaluate patients’ pain before the operation, 20 minutes after the puncture, 6 hours after the operation, and 1, 2, and 3 days after the operation.

The right jugular vein blood was collected before operation and at 6 h, 1 d, 2 d, and 3 d after operation. The levels of CRP (C-reactive protein) and neuron specific enolase (NSE) in plasma were detected by enzyme-linked immunosorbent assay (ELISA). Taking into account the influence of intraoperative bleeding and transfusion, the postoperative results were recorded, and corrected values were used, and the calculation was as follows:

2.7. Statistical Methods

The SPSS 21.0 software was used for statistical analysis. The area under the receiver operating characteristic (ROC) curve was used to evaluate the effectiveness of the model. The count data were described by frequency and constituent ratio, and the comparison between the radiomics-guided and general anesthesia groups was performed by chi-square test. Quantitative data are expressed as the . Shapiro–Wilk tests were performed to determine the normality of the data distributions. If the measurement data conform to the normal distribution, use -test to compare between above two groups.

3. Results

3.1. Diagnostic Efficacy of Radiomics Models

The selected models were evaluated, and the AUC of the model on the training set was 0.942 (95% CI 0.890-0.994), and the AUC of the model on the validation set was 0.941 (95% CI 0.885-0.996). The ROC curve is shown in Figure 2.

(a)

(b)

3.2. Comparison of General Information and Surgical Features

The basic information of patients is shown in Table 1. There was no significant difference in gender, age, and BMI between the radiomics-guided and general anesthesia groups (). All patients successfully completed hip surgery; postoperative showed promising results (Figures 3 and 4). The operation time of the radiomics-guided group was min, while that of general anesthesia group was min. There was no significant difference between the two groups (). The incidence of POCD in the radiomics-guided group was significantly lower than that in the general anesthesia group (5% vs. 30%, , ). The dosage of fentanyl in the radiomics-guided group was μg, and that in the general anesthesia group was μg, and the dosage of fentanyl in the radiomics-guided group was significantly less than that in the general anesthesia group ().

(a)

(b)

(a)

(b)

3.3. VAS Scores at Each Time Point

The VAS scores in the radiomics-guided and general anesthesia groups are shown in Figure 5. Compared with the general anesthesia group, the radiomics-guided group decreased significantly 20 minutes after puncture and 6 hours, 1 day, and 2 days after operation.

3.4. Serological Indexes at Each Time Point

The detailed comparison of CRP and NSE between the radiomics-guided and general anesthesia groups before and after the operation is shown in Table 2. Compared with that before operation, CRP and NSE increased significantly at each time point after operation, especially in the general anesthesia group.

4. Discussion

4.1. Analysis of POCD in Elderly Patients with Hip Surgery

The cognitive dysfunction is a common complication after hip arthroplasty. The incidence rate of POCD increases with age. The clinical incidence rate of elderly patients is 40% to 50% [24–26]. Anxiety, cognitive impairment, memory impairment, attention deficit, language, and social skill decline are common clinical manifestations, which seriously affect the quality of life of patients after. However, some scholars pointed out that surgical stress, operation time, physiological characteristics, anesthetic methods, extracorporeal circulation, anesthesia mode, residual anesthetic, and postoperative pain may be closely related to this complication [6–9]. Elderly patients still need general anesthesia because of their long-term special body position intolerance, difficulty in puncture in spinal canal, and use of anticoagulants and other factors, which may lead to postoperative cognitive dysfunction.

4.2. Advantages of Combined General Anesthesia and Iliac Fascia Block

Peripheral nerve block is often used as an important supplement and auxiliary means of general anesthesia [15–17]. The anesthetic effect is clear, which can shorten the recovery time, reduce the dosage of general anesthetic, and effectively reduce the incidence rate of related complications. Among various regional block techniques, iliac fascia space block is a simple and easy to master technique. Its mechanism is to inject local anesthetics into the space between the iliac fascia and the iliac muscle. Then, the femoral nerve and lateral femoral cutaneous nerve between them were blocked assisted by ultrasound-guided operation. This technology is used to treat the pain of patients with hip diseases, with rapid onset and exact effect. Some studies scored the pain of patients admitted to the emergency department [15, 27, 28]. It was found that the preoperative score was high, and the pain was not well controlled. However, immediate iliac fascia space block could effectively control the preoperative pain.

4.3. Comparison of Diagnostic Efficacy of Radiomics Models

The results showed that the AUC of ultrasonic image radiomics in the training set (AUC: 0.942, 95% CI 0.890-0.994) and the validation set (AUC: 0.941, 95% CI 0.885-0.996) is similar. It is proved that the ultrasonic image radiomics has great efficiency for the successful puncture.

4.4. Comparison of POCD Characteristics

The dosage of fentanyl in the radiomics-guided group was significantly lower than that in the general anesthesia group. The VAS score in the radiomics-guided group was significantly lower than that in the general anesthesia group at 6 h, 1 d, and 2 d after operation. In addition, the incidence of POCD in the radiomics-guided group (5%) was significantly lower than that in the general anesthesia group (30%). All these results indicated that the ultrasound radiomics-guided iliac fascia block can reduce the incidence of POCD after hip operation in elderly patients.

4.5. Comparison of CRP and NSE Levels

Some studies have shown that the central nervous system damage in patients with serum CRP and other inflammatory factors [29]. The CRP is an important cytokine with extensive biological activity in human body, which plays an important role in the central neuropathy progression. NSE is an enzyme widely existing in neurons, and its level changes are closely related to the degree of brain injury and prognosis of patients. Some studies have pointed out that inflammatory reaction and chronic oxidative stress reaction may lead to the synthesis and secretion of NSE [30, 31]. When it accumulates in the brain, it may lead to memory loss and progressive cognitive impairment. In this study, the CRP and NSE levels in both groups increased in a short period of time, indicating that proinflammatory factors are related to cognitive function. Compared with the general anesthesia group, the dosage of fentanyl in the radiomics-guided was lower, and the levels of CRP and NSE after operation were also lower. The results indicate that the ultrasound radiomics-guided iliac fascia block can reduce the use of general anesthesia drugs, reduce the release of inflammatory factors after and during operation, and improve postoperative cognitive function.

5. Conclusion

In conclusion, for elderly patients with hip surgery, the ultrasound radiomics-guided iliac fascia block can reduce the incidence of POCD and improve the effect of nerve block. However, the number of samples in this study is small, which needs to be further studied. The patients included in this study were insufficient and came from the same center. In addition, sufficient specimen is essential in ensuring the machine learning model’s efficiency.

Data Availability

The data used to support the findings of this study are available from the corresponding authors upon request.

Conflicts of Interest

All authors declare no conflicts of interest in this paper.

Authors’ Contributions

Jun Zha and Jinping Ni contributed equally to this work.

Acknowledgments

This work was supported by grants QNRC2016219 (to Dr. Qiao) from the Jiangsu Key Talent Youth Awards in Medicine, GSWS2019092 (to Dr. Qiao) from the Gusu Health Youth Talent Awards, 2020BYGT009 from the Open Project of Key Laboratory of Functional Genomics and Molecular Diagnosis of Gansu Province (to Dr. Zha), and 2019Y03 (to Dr. Zha) from the Young Talent Scientific Research Foundation of Suzhou Science & Technology Town Hospital.

References

R. Ehsani, S. Djalali Motlagh, B. Zaman, S. Sehat Kashani, and M. R. Ghodraty, “Effect of general versus spinal anesthesia on postoperative delirium and early cognitive dysfunction in elderly patients,” Anesthesiology and Pain Medicine, vol. 10, no. 4, article e101815, 2020.
View at: Publisher Site | Google Scholar
Y. X. Shen, Y. P. Chen, H. C. Zang, and G. Shao, “Evaluation of the efficiency of propofol versus isoflurane anesthesia interventions in treating elderly patients with postoperative cognitive dysfunction,” Medicine (Baltimore), vol. 99, no. 49, article e23511, 2020.
View at: Publisher Site | Google Scholar
W. Zhao, Y. Hu, H. Chen et al., “The effect and optimal dosage of dexmedetomidine plus sufentanil for postoperative analgesia in elderly patients with postoperative delirium and early postoperative cognitive dysfunction: a single-center, prospective, randomized, double-blind, controlled trial,” Frontiers in Neuroscience, vol. 14, article 549516, 2020.
View at: Google Scholar
Y. Gao, L. Liu, B. Zhao, Y. Wang, S. Yu, and H. Wang, “Effect of general and non-general anesthesia on postoperative cognitive dysfunction,” Journal of the College of Physicians and Surgeons–Pakistan, vol. 30, no. 4, pp. 407–411, 2020.
View at: Publisher Site | Google Scholar
U. Nilsson, K. Liander, O. Rooyackers, and L. I. Eriksson, “Patients’ experiences of early postoperative cognition and its relation to cognitive decline and inflammatory responses: a protocol for a mixed-methods study,” BMJ Open, vol. 9, no. 12, article e032650, 2019.
View at: Publisher Site | Google Scholar
B. Gao, B. Zhu, and C. Wu, “Preoperative serum 25-hydroxyvitamin D level, a risk factor for postoperative cognitive dysfunction in elderly subjects undergoing total joint arthroplasty,” The American Journal of the Medical Sciences, vol. 357, no. 1, pp. 37–42, 2019.
View at: Publisher Site | Google Scholar
S. Otomo, K. Maekawa, T. Baba, T. Goto, and T. Yamamoto, “Evaluation of the risk factors for neurological and neurocognitive impairment after selective cerebral perfusion in thoracic aortic surgery,” Journal of Anesthesia, vol. 34, no. 4, pp. 527–536, 2020.
View at: Publisher Site | Google Scholar
R. Yong and Y. Meng, “Preoperative neutrophil–lymphocyte ratio, an independent risk factor for postoperative cognitive dysfunction in elderly patients with gastric cancer,” Geriatrics & Gerontology International, vol. 20, no. 10, pp. 927–931, 2020.
View at: Publisher Site | Google Scholar
Y. Zhang, H. G. Bao, Y. L. Lv et al., “Risk factors for early postoperative cognitive dysfunction after colorectal surgery,” BMC Anesthesiology, vol. 19, no. 1, p. 6, 2019.
View at: Publisher Site | Google Scholar
C. Knaak, G. Vorderwülbecke, C. Spies et al., “C‐reactive protein for risk prediction of post-operative delirium and post-operative neurocognitive disorder,” Acta Anaesthesiologica Scandinavica, vol. 63, no. 10, pp. 1282–1289, 2019.
View at: Publisher Site | Google Scholar
A. Dizdarevic, F. Farah, J. Ding et al., “A comprehensive review of analgesia and pain modalities in hip fracture pathogenesis,” Current Pain and Headache Reports, vol. 23, no. 10, p. 72, 2019.
View at: Publisher Site | Google Scholar
S. N. Hofstede, M. G. J. Gademan, T. Stijnen, R. Nelissen, and P. J. Marang-van de Mheen, “The influence of preoperative determinants on quality of life, functioning and pain after total knee and hip replacement: a pooled analysis of dutch cohorts,” BMC Musculoskeletal Disorders, vol. 19, no. 1, p. 68, 2018.
View at: Publisher Site | Google Scholar
T. A. Lentz, S. Z. George, O. Manickas-Hill et al., “What general and pain-associated psychological distress phenotypes exist among patients with hip and knee osteoarthritis?” Clinical Orthopaedics and Related Research, vol. 478, no. 12, pp. 2768–2783, 2020.
View at: Publisher Site | Google Scholar
R. Shah, Y. F. Kuo, J. Westra, Y. L. Lin, and M. A. Raji, “Opioid use and pain control after total hip and knee arthroplasty in the us, 2014 to 2017,” JAMA Network Open, vol. 3, no. 7, article e2011972, 2020.
View at: Publisher Site | Google Scholar
C. Falyar and D. Tola, “Ultrasound-guided fascia iliaca blocks in the emergency department,” Geriatric Nursing, vol. 40, no. 4, pp. 441–444, 2019.
View at: Publisher Site | Google Scholar
J. L. Glomset, E. Kim, J. M. Tokish et al., “Reduction of postoperative hip arthroscopy pain with an ultrasound-guided fascia iliaca block: a prospective randomized controlled trial,” The American Journal of Sports Medicine, vol. 48, no. 3, pp. 682–688, 2020.
View at: Publisher Site | Google Scholar
H. Wang, Q. Li, and Y. Ni, “Ultrasound-guided “hourglass-pattern” fascia iliac block combined with sacral plexus and gluteal epithelial nerve block for an elderly hip fracture patient with organ failure,” Medicine (Baltimore), vol. 99, no. 25, article e19732, 2020.
View at: Publisher Site | Google Scholar
Z. Han, K. Liu, Z. Li, and P. Luo, “A pre-check operator for reducing algorithmic optimisation time in image processing applications,” Enterprise Information Systems, vol. 13, pp. 1–13, 2021.
View at: Publisher Site | Google Scholar
D. Hua, Y. Wang, S. Ji, J. Zha, and Y. Xia, “The multi-objective optimization of combustion system operations based on deep data-driven models,” Journal of Medical Imaging and Health Informatics, vol. 9, no. 5, pp. 951–955, 2019.
View at: Google Scholar
H. Cui, D. Zhang, F. Peng et al., “Identifying ultrasound features of positive expression of Ki67 and P53 in breast cancer using radiomics,” Asia-Pacific Journal of Clinical Oncology, vol. 17, no. 5, pp. e176–e184, 2021.
View at: Publisher Site | Google Scholar
Y. Guo, Y. Hu, M. Qiao et al., “Radiomics analysis on ultrasound for prediction of biologic behavior in breast invasive ductal carcinoma,” Clinical Breast Cancer, vol. 18, no. 3, pp. e335–e344, 2018.
View at: Publisher Site | Google Scholar
Y. Liu, Q. Fang, A. Jiang, Q. Meng, G. Pang, and X. Deng, “Texture analysis based on u-net neural network for intracranial hemorrhage identification predicts early enlargement,” Computer Methods and Programs in Biomedicine, vol. 206, 2021.
View at: Publisher Site | Google Scholar
Y. Antonacci, J. Toppi, D. Mattia, A. Pietrabissa, and L. Astolfi, “Single-trial connectivity estimation through the least absolute shrinkage and selection operator,” in 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), vol. 2019, pp. 6422–6425, Berlin, Germany, 2019.
View at: Publisher Site | Google Scholar
S. Tang, W. Huang, K. Zhang, W. Chen, and T. Xie, “Comparison of effects of propofol versus sevoflurane for patients undergoing cardiopulmonary bypass cardiac surgery,” Pakistan Journal of Medical Sciences, vol. 35, no. 4, pp. 1072–1075, 2019.
View at: Publisher Site | Google Scholar
Q. X. Xiao, Q. Liu, R. Deng, Z. W. Gao, and Y. Zhang, “Postoperative cognitive dysfunction in elderly patients undergoing hip arthroplasty,” Psychogeriatrics, vol. 20, no. 4, pp. 501–509, 2020.
View at: Publisher Site | Google Scholar
W. Yan, H. Mao, and P. Qiu, “Effects of different analgesia regimens on early post-operative cognitive dysfunction in elderly patients undergoing radical resection of cervical carcinoma,” Experimental and Therapeutic Medicine, vol. 18, no. 2, pp. 1465–1469, 2019.
View at: Publisher Site | Google Scholar
E. Dochez, G. J. van Geffen, J. Bruhn, N. Hoogerwerf, H. van de Pas, and G. Scheffer, “Prehospital administered fascia iliaca compartment block by emergency medical service nurses, a feasibility study,” Scandinavian journal of trauma, resuscitation and emergency medicine, vol. 22, no. 1, p. 38, 2014.
View at: Publisher Site | Google Scholar
L. Haines, E. Dickman, S. Ayvazyan et al., “Ultrasound-guided fascia iliaca compartment block for hip fractures in the emergency department,” The Journal of Emergency Medicine, vol. 43, no. 4, pp. 692–697, 2012.
View at: Publisher Site | Google Scholar
X. Liu, Y. Yu, and S. Zhu, “Inflammatory markers in postoperative delirium (POD) and cognitive dysfunction (POCD): a meta-analysis of observational studies,” PLoS One, vol. 13, no. 4, article e0195659, 2018.
View at: Publisher Site | Google Scholar
T. Luescher, J. Mueller, C. Isenschmid et al., “Neuron-specific enolase (NSE) improves clinical risk scores for prediction of neurological outcome and death in cardiac arrest patients: results from a prospective trial,” Resuscitation, vol. 142, pp. 50–60, 2019.
View at: Publisher Site | Google Scholar
D. W. Park, S. H. Park, and S. K. Hwang, “Serial measurement of S100B and NSE in pediatric traumatic brain injury,” Child's Nervous System, vol. 35, no. 2, pp. 343–348, 2019.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2022 Jun Zha et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

250

Downloads

528

Citations