Passive leg raising modulates low-frequency oscillation propagation in peripheral atherosclerosis: A pilot study

Abstract

Objective

Low-frequency oscillations (LFOs) observed in the periphery may reflect physiological processes. The aim of this study was to investigate these processes’ effects on LFOs and the differences between healthy subjects and those with peripheral arteriosclerosis disease (PAD).

Methods

14 PAD patients and 25 healthy controls were studied in resting (RS) and passive leg raising (PLR) states. We simultaneously measured LFOs at the peripheral left earlobes (LE), right earlobes (RE), left fingertips (LF), right fingertips (RF), left toes (LT), and right toes (RT), along with coherence and phase shift analysis processing.

Results

The coherence coefficients in the PAD group were lower than those in the healthy group (p < .01), and the phase shifts in the PAD group were higher than those in the healthy group (p < .01) in a resting state. Mild to moderate PAD patients had greater coherence coefficients and smaller phase shifts than severe PAD patients. 0.05 Hz PLR LFOs originating in the LT can be observed in other peripheral positions. The proportion of occurrence times for 0.05 Hz PLR LFOs peaks observed at different peripheral positions was different in healthy subjects, patients with bilateral multiple lower limb arteriosclerosis, and those with left or right lower limb arteriosclerosis.

Conclusion

The coherence coefficient and phase shift characteristics of LFOs were different between healthy subjects and PAD patients. LFOs have the potential to provide valuable physiological process information associated with atherosclerosis in the periphery.

Keywords

Low-frequency oscillations peripheral atherosclerosis passive leg raising

Introduction

Peripheral Atherosclerosis Disease (PAD) is rarely a direct cause of death, and therefore, it receives insufficient attention in the medical literature. However, PAD patients have elevated risk of cardiovascular and cerebrovascular events, such as coronary artery death, myocardial infarction, stroke, and severe atherosclerotic disease, according to the Reduction of Atherothrombosis for Continued Health (REACH) study.¹ 90% of PAD patients who undergo coronary angiography have coronary artery disease, while 50% have cerebrovascular disease.² These findings imply that PAD may be associated with systemic vascular disease. In an asymptomatic middle-aged cohort, the prevalence of femoral atherosclerosis was even higher than that of the carotid artery.³ Whether symptomatic or asymptomatic, PAD is a risk factor for both cardiovascular and cerebrovascular events.⁴ Symptomatic PAD has been identified as a window indication of systemic vascular atherosclerosis in coronary artery disease and in other risk states.⁵ Therefore, investigating changes in blood circulation and physiologic processes in PAD may be useful in identifying potential cardiovascular and cerebrovascular events.

Currently, computed tomography angiography and ultrasonic imaging are the primary means of PAD clinical diagnosis. These methods usually require advance appointments in specific hospital rooms with particular environments, which is not conducive to timely PAD discovery. Therefore, a new method/technique is needed to enable the detection of arteriosclerosis at any time bedside. Low-frequency oscillations (LFOs) in the range of 0.01 Hz–0.15 Hz are slow, spontaneous variations in hemodynamic parameters such as oxygenation and pressure.^6–8 In general, they can be seen in the brain and the periphery.⁹ Studies have shown that LFOs are related to the autonomic physiological process and may be used to reflect changes in blood circulation.¹⁰ LFOs in the symmetric periphery have high correlation coefficients and near-zero time delays, and may be used as biomarkers to reflect vascular integrity.¹¹

Based on the above results, researchers have further explored the propagation characteristics of LFOs in stroke patients.¹² The results show that the correlation coefficients of symmetric peripheral LFOs in stroke patients are significantly reduced. This may indicate that vascular stenosis in stroke patients can block or delay the propagation of LFOs, resulting in reduced correlation and increased time delay.¹² These results demonstrate that LFOs can reflect the differences in blood circulation between stroke and healthy individuals. However, these differences in peripheral LFOs between healthy and stroke patients may be due to concurrent peripheral arteriosclerosis, rather than stroke. The physiological mechanism of using LFOs in the periphery to reflect ischemic features in the brain is difficult to elucidate. Therefore, we hypothesized that these LFOs in the periphery are better suited to reflect peripheral blood circulation than those in the brain. As such, they may serve as an effective and accurate tool for assessing peripheral blood circulation.

The purpose of this study was to investigate the propagation characteristics of LFOs in PAD patients, to identify the effects of physiological processes on LFOs, and the differences between healthy subjects and PAD patients. The PLR experiment was designed to generate exogenous biomarkers to modulate LFOs in healthy subjects and PAD patients. The main contributions of this paper are as follows: (1) we investigated the characteristic differences in LFOs between a healthy group and a PAD group, and clarified the characteristic differences in LFOs between mild to moderate PAD patients and severe PAD patients, confirming that LFOs have the potential to provide valuable patient information associated with peripheral atherosclerosis. (2) We clarified that the differences in the proportion of 0.05 Hz PLR LFO peak occurrence times observed at different peripheral positions was different in healthy subjects, patients with bilateral multiple lower limb arteriosclerosis, and those with left or right lower limb arteriosclerosis, confirming that LFOs may serve as perfusion biomarkers to indicate the presence of peripheral atherosclerosis.

Materials and methods

Participants

Ethical approval: This study was approved by Rocket Force Characteristic Medical Center Committee (Ref: KY2021036), and conformed to the standards set by the Declaration of Helsinki (latest revision: 59th WMA General Assembly). The principle of signal detection in this study is NIRS-based oxygen saturation detection with non-invasive and short duration, without any risk to patients. Each patient or guardian received a detailed explanation of the above characteristics before each experiment. All subjects were recruited on a voluntary basis and were enrolled in this study with the consent of either the patients or their guardians. There was no specific monetary reward for participating. However, before participating in the study, we verbally thanked the patients and/or guardians for their contributions to the study.

The study included a total of 39 volunteers (14 PAD patients, 44 to 90 y.o., and 25 healthy controls, 21 to 59 y.o.). The exclusion criteria were acute coronary syndrome, heart failure, active infectious or autoimmune disease, limb fractures, and chronic liver or kidney disease. Three of the PAD patients did not have detailed diagnostic information and were not included in the data analysis. Experienced clinicians and their teams were invited to discuss the clinical diagnosis results of all patients in this study. Less than 70% vascular stenoses was considered mild to moderate (5 of the 11 PAD patients), while greater than 70% was considered severe (6 of the 11 PAD patients). The patients were divided into three categories based on diagnostic results: bilateral multiple lower limb arteriosclerosis (6 of the 11 PAD patients), left lower limb arteriosclerosis (3 of the 11 PAD patients), and right lower limb arteriosclerosis (2 of the 11 PAD patients). This study was a preliminary study consisting of qualitative and semi-quantitative analysis. DSA images were not available for all patients; some had CTA and some had arterial color ultrasound. The corresponding diagnostic images for all patients are shown in Figure 8 (Appendix A). All diagnostic findings were collected in a correspondingly standardized manner with an experienced physician who was blinded to the subjects’ analysis results. Basic information for all patients is shown in Table 2 and Table 3 (Appendix A).

Procedure

Two experimental states were used in this study: the resting state (RS) and the passive leg raising state (PLR), as shown in Figure 1(a) and (b). Our team independently developed a multi-channel LFO detection device (MNO) based on NIRS to simultaneously detect the LFOs at multiple positions in the periphery,¹³ as shown in Figure 1(c). The sampling rate of 31.25 Hz allows one to fully sample cardiac, respiratory, and LFO waveforms without aliasing.¹³ The device collects the light intensity change signal after reflection and absorption by the body tissue. We used the modified Beer–Lambert law (MBLL) to translate the light intensity signals into oxygenated (∆[HbO]) and deoxygenated (∆[Hb]) hemoglobin concentrations.^14,15 Finally, we extracted LFOs in the frequency range of 0.01 Hz–0.15 Hz through a zero-phase Butterworth bandpass filter.¹⁶ The signal detection positions in different states are shown in Figure 1(d).

Figure 1.

The data acquisition process on peripheral arteriosclerosis patients in resting state (a) and passive leg raising state (b), the multi-channel LFO detection device (c) and the signal detection positions in different states (d).

Before the experiments, professional medical staff brought the subjects to the designated experimental room. Each participant’s name (due to privacy concerns, the name is replaced with an ID number), age, weight, height, systolic and diastolic blood pressures, hyperglycemia, and hyperlipidemia were recorded, as shown in Table 2. The participants were in a room with dim light, an indoor temperature about 25°C, relative humidity of 30%–40%, and air pressure about 100k Pa. A resting state refers to the state in which an individual is awake and relaxed, and no task or activity is required. In the RS experiment, the subjects were in a supine position with their eyes open on a comfortable bed. According to the range of global circulation times,¹⁷ the signal acquisition time must far exceed the blood circulation time. Therefore, the experiment would last about 12 min, and we would delete any unstable data at either end of the data segment to increase overall stability and retain 10 min of data from six peripheral positions, as shown in Figure 2(a).

Figure 2.

Resting state (a) and passive leg raising state (b) experiments.

In the PLR experiment, we used gravity to pump blood from the legs to the heart. About 300 mL of blood can be drawn from the lower extremities with each passive leg rising.¹⁸ Jabot et al.¹⁹ investigated the hemodynamic effects of PLR initiation in the semi-sitting and supine positions. Their results showed that initiating PLR in the semi-sitting position mobilized an additional 150 mL of blood compared to the supine position. However, in our study, due to the limitations in the PAD patients’ physical states, we initiated PLR in a supine position. Medical professionals raised the subjects’ left legs at 0.05 Hz in the following order: (1) raise the left leg at an angle with the ground of approximately 45° (duration ∼5 s), (2) lower the leg onto the bed for ∼5 s, and (3) rest for 9 s. After 2 min, the subject had 1 min of rest, and then the cycle was repeated. Three raising-resting blocks were used in the experiment, as shown in Figure 2(b). No conscious innervation was required for any of the procedure.

Data analysis methods

We applied two methods to mitigate the effect of motion artifacts—cubic-spline interpolation²⁰ and kurtosis-based wavelet filter.²¹ The former eliminates abnormal “spikes” while the latter relieves unexpected sudden movements (e.g., sweeping gestures and sudden head movements). We detrended the 10 min timecourse dataset (∼18,751 data points at a sampling rate fs = ∼31.25 Hz) for each parameter under a specific physiological condition to remove any baseline shifts. We divided detrended data into 11 segments with 90% overlap in two adjacent segments, resulting in a length of ∼9375 (300*31.25) data points for each segment. After applying a Hanning window to reduce the effect of spectral leakage, we determined each segment’s frequency domain spectrum by fast Fourier transform. Then, we extracted the LFOs’ coherence coefficients²² and phase shifts²³ between both channels. Next, we weighted and averaged the coherence coefficients and phase shifts of 11 segments’ LFOs to reduce the error score. Rank-sum test (i.e., ranksum) was used to explore the differences between the PAD patients and the healthy controls.

To reduce the false discovery rate (FDR) for positive results, we used the Benjamini–Hochberg method to correct the results, keeping the false-positive probability below 0.05. The analyses was focused on oxygenated hemoglobin (Δ[HbO]), which has a higher signal-noise ratio (SNR) than deoxygenated hemoglobin (Δ[Hb]).^24,25 However, it is also possible that the veins (which contain more [Hb]) may have been less reactive to blood pressure variation than the arteries (which contain more [HbO]), but Δ[Hb] demonstrated the least magnitude and SNR in the LFOs. Therefore, we excluded the Δ[Hb] from the following discussion. A flowchart of LFO extraction and processing and analysis methodology is shown in Figure 3.

Figure 3.

Flowchart of LFO extraction and processing and data analysis methodology.

Results

The characteristics of LFOs in PAD patients under different states

In the resting state, the coherence coefficients in healthy subjects’ symmetrical earlobes (LE-RE), fingertips (LF-RF), and toes (LT-RT) were greater than those in the PAD patients (LE-RE: p < .001, LF-RF: p < .001, LT-RT: p < .001). Also, the phase shifts in the healthy subjects were significantly smaller than those in the PAD patients (LE-RE: p < .001, LF-RF: p < .01, LT-RT: p < .001), as shown in Figure 4 (a) and (b). The PLR was found to attenuate the coherence coefficients of the symmetrical earlobes, fingers, and toes (LE-RE: p < .001, LF-RF: p < .001, LF-RF: p < .001) and increase the corresponding phase shifts (LE-RE: p < .001, LF-RF: p < .001, LF-RF: p < .001) in healthy subjects. PLR did not significantly change the LFOs’ coherence coefficients or phase shifts in PAD patients. These findings indicate that peripheral arteriosclerosis may limit the PLR-perturbing effect.

Figure 4.

Statistical results for coherence coefficients (a) and phase shifts (b) in symmetric positions, in both resting and PLR states. (p < .001 (***), p < .01 (**), p < .05 (*))

The characteristics of LFOs in PAD patients with semi-quantified degree

Further, we explored the LT-RT coherence coefficients and LT-RT phase shifts in patients with mild and moderate PAD and patients with severe PAD, as shown in Figure 5. The results show that the coherence coefficients were lower among the severe PAD than the mild and moderate PAD in the resting state; the average difference was 0.2. PLR appears to have a greater effect on coherence coefficients in healthy controls and patients with mild and moderate PAD, and less of an effect in patients with severe PAD, as shown in Figure 5(a). In Figure 5(b), patients with severe PAD had larger phase shifts than those with mild or moderate PAD in the resting state. It should be noted that Patients #14, #2, and #8 produced opposite phase shifts between the resting state and the PLR state. This may have been due to PLR accelerating the upward perfusion of blood flow in the left leg, resulting in a wide disparity in blood flow velocity between the left and right legs.

Figure 5.

Correlation coefficients (a) and phase shifts (b) in patients with mild and moderate PAD and patients with severe PAD, in resting and PLR states, and statistical results for the healthy control groups.

PLR LFO propagation in PAD patients of different categories

We further investigated whether LT PLR LFOs peaks (PLR LFOs’ source) could be observed in the peripheral sites of healthy controls and PAD patients. Figure 6(a) shows the propagations of 0.05 Hz PLR LFOs in healthy subjects, as well as those in patients with left or right lower limb arteriosclerosis. In Figure 6(b), the results show that 0.05 Hz PLR LFOs peaks were not only in the LT, but also in LE and RE, LF, RF, and RT in the healthy subjects. This finding indicated that 0.05 Hz PLR LFOs peaks may serve as “labels” along with blood flow, and spread through vascular channels to other peripheral sites. In patients with left lower limb arteriosclerosis, 0.05 Hz PLR LFOs peaks are observable only in the RE and RT, as shown in Figure 6(c). This finding indicated that peripheral arteriosclerosis in the left leg may block the flow of 0.05 Hz PLR LFOs from the left foot (PLR source) to other positions. In patients with right lower limb arteriosclerosis, 0.05 Hz PLR LFOs peaks in the LE, RE, LF, RF, and LT, but not in the RT, as shown in Figure 6(d). This finding indicated that peripheral arteriosclerosis of the right leg may prevent the spread of LFOs from the LT to the RT. These results suggested that the propagation of PLR LFOs may be related to the peripheral arteriosclerosis.

Figure 6.

The propagation of PLR peaks in healthy subjects and patients with lower limb arteriosclerosis (a), and the power spectra of PLR LFOs in healthy subjects (b), and patients with left (c) or right (d) lower limb arteriosclerosis.

The proportions of 0.05 Hz PLR LFO occurrence times at different peripheral positions were also counted. Each subject had three PLR periods (60 s −180 s, 240 s −360 s, and 420s −540 s) within 600 s. There were 25 healthy subjects and 11 PAD patients, including three patients with left lower limb arteriosclerosis, two patients with right lower limb arteriosclerosis, and six patients with bilateral multiple lower limb arteriosclerosis. The number of 0.05 Hz PLR LFOs observed at different peripheral positions during PLR among healthy subjects and PAD patients is shown in Table 1. We found that there were significantly more 0.05 Hz PLR peaks in healthy individuals than in PAD patients. 0.05 Hz PLR LFOs peaks were difficult to observe in the peripheral positions outside the LT in the patients with left lower limb arteriosclerosis and bilateral multiple lower limb arteriosclerosis. 0.05 Hz PLR LFOs peaks were difficult to observe in RT in the patients with right lower limb arteriosclerosis. These results indicated LFOs may serve as perfusion biomarkers to indicate the presence of peripheral atherosclerosis.

Table 1.

0.05 Hz PLR LFOs peaks were counted at different peripheral positions during passive leg rising in 25 healthy subjects and 11 PAD patients.

Positions	LE (%)	RE (%)	LF (%)	RF (%)	LT (%)	RT (%)
Healthy subjects (25N*3 period)	89	72	81	91	100	100
PAD patients (11N*3 period)	39	30	42	27	94	18
Left lower limb arteriosclerosis (3N*3 period)	33	11	22	11	89	11
Right lower limb arteriosclerosis (2N*3 period)	78	78	67	56	100	22
Bilateral multiple lower limb arteriosclerosis (6N*3 period)	20	13	40	20	93	20

Comparative analysis of the ankle-brachial index with PAD patients

We collected the ABI of Patients #11 and #14 before and after vascular surgery for comparative analysis with LFOs. Patient #11 underwent left femoral artery thrombectomy and balloon dilation. Patient #14 underwent balloon dilatation of the right lower limb artery. To rule out the effects of different surgical types, we further collected ankle-brachial index (ABI) scores from Patient #10 (there was little difference in ABI between the right lower limb and the left lower limb) and Patient #13 (there was a significant difference in ABI between the right lower limb and the left lower limb) before vascular surgery, as shown in Figure 7(a). The changes in LFO characteristics in the left and right lower limbs before surgery were compared between Patient #10 and Patient #13; the influence of the type of surgery did not have to be considered. In Figure 7(b), the results showed that the LT-RT coherence coefficients of all PAD patients in a resting state were significantly lower than the mean value for all the healthy subjects. This indicated that the LFOs’ LT-RT coherence coefficients could distinguish healthy subjects from PAD patients. There was a significant reduction in the ABI difference between the left and right legs in Patient #11 and Patient #14 after surgery, as well as the LFOs’ phase shifts. The change in LFO phase shifts was consistent with the change trend for patients’ ABI, indicating that LFO phase shift characteristics may indicate the presence of peripheral atherosclerosis. The comparison between Patient #10 and Patient #13 showed that the difference in ABI index between the left and right legs of Patient #13 was significantly greater than that of Patient #10. Also, the LFO phase shifts between the left and right legs of Patient #13 were also significantly greater than those of Patient #10, indicating again that LFO phase shift characteristics may reflect the differences between the left and right legs of PAD patients, as shown in Figure 7(b).

Figure 7.

Ankle-brachial index (a), the coherence coefficients and phase shifts of LFOs in the resting state (b), and the power spectral density of LFOs in the PLR state (c).

No significant 0.05 Hz PLR LFOs were observed in RT, in either Patient #11 or Patient #14, before vascular surgery. However, 0.05 Hz PLR LFOs were observed in RT in both patients after vascular surgery, indicating that the patients’ blood circulation had been partially restored after treatment, as shown in Figure 7(c). The amplitude of the RT 0.05 Hz PLR LFOs peaks in Patient #13 was significantly smaller than that in Patient #10, which may indicate that Patient #13 had more severe atherosclerosis in the lower limb vessels and had a greater effect on blocking or delaying LFOs. These findings corresponded to those of Patient #10 with lower ABI in the right lower limb.

Discussion

To the best of our knowledge, this is the first study assessing the relationships between the changes in LFOs and the presence of peripheral atherosclerosis disease. To explore the propagation characteristics of LFOs in patients with PAD, we revealed the LFOs’ changes after disruption by arteriosclerosis in a resting state. The reduced coherence coefficients with increased LFO phase shifts were observed in the symmetric periphery of PAD patients. Additionally, PLR decreased LFOs’ symmetry in healthy controls, but had little effect on LFOs’ symmetry in PAD patients. We hypothesized that this finding may indicate that vascular stenosis/injuries limit the PLR-perturbing effect. Therefore, we further explored the differences in propagation of PLR LFO in healthy subjects, as well as in patients with left or right lower limb arteriosclerosis. The results suggested that the PLR LFOs’ propagation may be related to the lower limb arteriosclerosis.

An additional observation derived from the current results is that because the LFOs are derived by the application of a bandpass filter (0.01 to 0.15 Hz), the heartbeat (∼1 Hz) and respiratory (∼0.2 Hz) signals are not the aliasing of these signals into the LFO signal. Therefore, the LFOs that we have identified in the periphery are independent of the fluctuations arising from the cardiac pulsation (measured by pulse oximeter) as well as respiration (measured by respiration belt). This provides strong counterevidence that the non-neuronal LFO in BOLD is primarily the aliased signal arising from cardiac pulsation and respiration. However, this idea does not preclude the possibility that the LFO represents physiologic processes indirectly related to cardiac pulsation and respiration. Therefore, we excluded cardiopulmonary heart disease patients from the PAD subjects to the greatest extent possible to ensure that any asymmetric changes in LFO were primarily due to arteriosclerosis.

In the heart, LFO signals in the blood must be in phase, as all of the blood is mixed in the left ventricle before being pumped throughout the body.¹⁰ When the blood exits the heart at the ascending aorta, pathways diverge for blood traveling to different locations throughout the body. Because the body’s vasculature is relatively symmetrical, the LFO signals recorded in the two earlobes, fingertips, and toes are more correlated than those in the finger-toe or finger-brain pairs, and they also have smaller relative phase shifts. However, peripheral arteriosclerosis increases the vascular system’s complexity, resulting in differences in the length, diameter, and elasticity of symmetrical vascular pathways, which together may render the blood signal less uniform as it travels.

Our previous work has focused on the transmission characteristics of systemic LFOs in the resting state. It has revealed that LFOs may serve as an endogenous source of contrast, and spread symmetrically in healthy subjects.¹⁴ Symmetry differences in LFOs can be used to assess vascular injury diseases. However, because of the interaction/feedback effects between neuromodulation and non-neuromodulation, explaining the physiological origin of these symmetry differences is difficult. In this study, fixed-frequency (0.05 Hz) PLR signals were added to a specific location, which is the equivalent of adding a “label” to the LFO. Based on the principle of PLR to mobilize upper limb blood perfusion, we hypothesized that PLR LFOs may serve as exogenous biomarkers to chart the overall course of blood flow and to reflect changes in the physiologic processes associated with hypo-perfusion.

The development of peripheral LFOs for peripheral arterial disease research offers the following advantages. First, LFOs based on changes in oxygenation levels are available through near infrared spectroscopy devices, which are typically portable and simple to use, as shown in Figure 1(c). This can be used to detect arteriosclerosis at any time at bedside, which will aid in future clinical applications or possibly facilitate at-home PAD detection. Second, the function of these LFOs may be used to reflect systemic physiological circulatory effects in PAD patients. Therefore, LFOs might not only be suitable for clinical application of PAD, but also for other diseases related to peripheral blood circulation in the future. Third, exploring these LFOs may be conducive to revealing the interaction/influence between peripheral blood flow/physiological processes, enabling early detection and treatment of potential atherosclerosis events, before hospitalization.

This is only a preliminary study, so it has the following limitations. First, the physiological tests and angiographic study were not comprehensive. Among the 14 PAD patients, only 6 CTA examinations and 5 vascular ultrasound Doppler examinations were included, as shown in Figure 8. In the future, we will use DSA as the gold standard to further analyze the morphological description of each patient’s vascular tree. Second, data from only 14 PAD patients were collected due to the COVID-19 epidemic in China. It would have been difficult to collect data from a large number of PAD patients in such a short period of time. Due to the number of PAD patients, we did not carry out more detailed classifications of the degree of disease. The analysis of the degree of disease in PAD patients is based on semi-quantified mild to moderate and severe. In the future, we will further expand the sample size and conduct grouping studies on the symptoms and severity of the disease to further optimize the present study’s conclusions. Of course, we will also consider different states, postures, and behaviors’ effects on LFOs in future studies, including the characteristics of LFOs during sleep, anesthesia, deep coma, and standing, sitting, supine, and semi-lying positions. The MNO-based LFO detection device is only 5.2*5.2*17.8 cm³ (as shown in Figure 1(c)), and can be easily fixed to the arm or waist. Therefore, we will also detect the propagation characteristics of LFOs in static walking, jumping, and other motion states in the future. Combined with their portability and accessibility, LFOs may offer clinicians an additional tool to incorporate into their practice to improve patient prognostication, treatment allocation, and therapeutic response assessment.

Conclusion

Detecting peripheral LFOs in PAD patients in RS and PLR states, we found symmetrical differences in circulating LFOs in both healthy individuals and PAD patients. LFOs may be used as biomarkers to reflect the valuable patient information about physiologic processes that are associated with peripheral atherosclerosis. Exploring these LFOs may be conducive to revealing the interaction/influence between peripheral blood flow/physiological processes. Unlike the existing PAD diagnosis methods, LFOs present no restrictions on the operating environment and are more flexible. Therefore, LFOs may be used to enable arteriosclerosis detection at any time at bedside in the future, which would be conducive to timely PAD discovery. However, we have to admit that the practicability of LFOs remains unclear. Further studies are needed to confirm the accuracy, sensitivity, and specificity of this method.

Data Availability Statement

Data underlying the results presented in this paper are not publicly available at this time, but may be obtained from the authors upon reasonable request.

Ethical statement

Informed consent

The code generated and analyzed in this study is available from the corresponding author upon reasonable request.

Footnotes

Acknowledgments

We would like to thank Dr Bian Ce, Chief of the Vascular Surgery Department at Rocket Army Specialty Medical Center, for providing several cases.

Author contributions

Yunfei Ma contributed reagents, materials, analysis tools and data, and wrote the manuscript; Kexin Luo and Zhengxuan Zhou analyzed and interpreted the data; Xiaoli Li performed the experiments; Shimin Yin and Yingwei Li conceived and designed the experiments.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by the National Natural Science Foundation of China 61827811, and in part by the Hebei Provincial Funding Project for Overseas Students C20200364.

Institutional Review Board Statement

This study was conducted in accordance with the Declaration of Helsinki, and approved by the ethics committee at PLA Rocket Force Medical Center (KY2021036).

ORCID iD

Yunfei Ma

Appendix

We have provided details and corresponding clinical imaging for all PAD patients. Details on the 11 PAD patients are shown in Tables 2 and 3. Diagnostic imagery for the 11 PAD patients is shown in Figure 8. Table 2.

Details about the 11 PAD patients.

ID	Sex	Age	Height (m)	Weight (kg)	Body temperature (°C)	Pulse rate	Respiratory rate	Systolic pressure (mmHg)	Diastolic pressure (mmHg)	Hyperglycemia	Hyperlipidemia
Patient #2	Male	51	1.72	70	36.1	67 times/min	23 times/min	123	85	Yes	Yes
Patient #3	Female	81	1.65	65	36.2	68 times/min	27 times/min	185	96	Yes	No
Patient #4	Female	79	1.60	50	36.8	77 times/min	25 times/min	166	84	Yes	No
Patient #7	Female	57	1.67	70	36.4	78 times/min	18 times/min	163	87	Yes	No
Patient #8	Male	59	1.74	72	36.2	77 times/min	18 times/min	119	80	Yes	No
Patient #9	Female	62	1.53	57	36.5	81 times/min	18 times/min	158	95	Yes	No
Patient #10	Female	78	1.62	60	36.8	93 times/min	17 times/min	137	88	Yes	No
Patient #11	Male	70	1.69	65	36.7	86 times/min	24 times/min	160	80	Yes	No
Patient #12	Female	88	1.72	75	36.5	70 times/min	23 times/min	123	85	Yes	Yes
Patient #13	Male	79	1.75	80	36.4	72 times/min	21 times/min	134	93	Yes	No
Patient #14	Male	75	1.74	80	36.2	81 times/min	29 times/min	160	91	Yes	No

Table 3.

Details about the 11 PAD patients.

ID	Intermittent claudication	Traumatic critical limb ischemia	Clinical diagnosis
Patient #2	Yes	No	1. Occlusion of the left posterior tibial artery, distal popliteal artery, anterior tibial artery and peroneal artery
Patient #2	Yes	No	2. Spot calcification on the wall of the right common iliac artery.
Patient #3	No	No	1. Intima-media of the femoris, popliteal, anterior tibial, posterior tibial, and peroneal arteries of the right lower limb are not uniformly thickened and are rougher; considerable uneven and strong echoic plaque is apparent on the wall.
Patient #4	No	No	1. Local stenosis of the left popliteal artery lumen, with an area stenosis rate of about 60%.
Patient #7	No	No	1. Left common femoral artery stenosis area 63%
Patient #7	No	No	2. The area of the plaque in the left common femoral artery was 23.3 mm*6.3 mm.
Patient #8	No	No	1. Considerable atherosclerotic plaque in the right lower limb with no arterial stenosis.
Patient #9	No	No	1. Incomplete occlusion in the left inferior femoral artery, anterior tibial, posterior tibial, and peroneal arteries.
Patient #10	Yes	No	1. The left anterior tibial, posterior and middle-distal peroneal lumen were not explored, and the possibility of occlusion could not be ruled out.
Patient #10	Yes	No	2. Abnormal right popliteal, anterior tibial, and posterior tibial fibular arteries were viewed, and there was incomplete occlusion.
Patient #11	Yes	No	1. There was segmental stenosis and occlusion in the abnormal left femoral, popliteal and posterior tibial arteries.
			2. Severe stenosis of distal lumen in the right posterior tibial artery; possible occlusion.
			3. Potential bilateral anterior tibial artery lumen double-occlusion possible; potential bilateral dorsal pedis artery lumen double-occlusion
Patient #12	Yes	No	1. Intima-media of the left and right femoral, popliteal, anterior tibial, posterior tibial and peroneal arteries are unevenly thickened and rougher, and considerable uneven and thick echoic plaque can be seen on the wall.
Patient #13	Yes	No	1. Diffuse mixed plaque was found in most of the internal iliac arteries and distal segments of the external iliac arteries, and the lumen were severely narrowed to occlusion; bilateral anterior tibial artery and right peroneal artery lumen occlusion
Patient #14	Yes	No	1. Bilateral internal iliac artery; mild to moderate stenosis on the left popliteal artery lumen
Patient #14	Yes	No	2. Mild stenosis of the lumen in the bilateral common iliac artery, left external iliac artery, bilateral anterior tibial artery and right femoral artery.

Figure 8.

Diagnostic imagery of the 11 PAD patients.

References

Cacoub

Abola

Baumgartner

, et al. Cardiovascular risk factor control and outcomes in peripheral artery disease patients in the Reduction of Atherothrombosis for Continued Health(REACH) Registry. Atherosclerosis 2008; 204(2): e86.

Stehouwer

Clement

Davidson

, et al. Peripheral arterial disease: a growing problem for the internist. Eur J Intern Med 2009; 20(2): 132–138.

Strangman

Franceschini

Boas

. Factors affecting the accuracy of near-infrared spectroscopy concentration calculations for focal changes in oxygenation parameters. Neuroimage 2003; 18(4): 865.

Perry

Bridges

Zhu

, et al. Getting to the root of fine motor skill performance in dentistry: brain activity during dental tasks in a virtual reality haptic simulation. J Med Internet Res 2017; 19(12): e371.

Cleeman

. Executise summary of third report of the ntiual cholesterol education program (NCEP) expert panel on section evaluation, and treatment of high blood cholesterol in adults atiult treatment panel I. JAMA 2001; 285: 2486–2489.

Buzsaki

Draguhn

. Neuronal oscillations in cortical networks. Science 2004; 304(5679): 1926–1929.

Katura

. Quantitative evaluation of interrelations between spontaneous low-frequency oscillations in cerebral hemodynamics and systemic cardiovascular dynamics. Neuroimage 2006; 31(4): 1592–1600.

Ming

Xin

, et al. Correlation analysis between prefrontal oxygenation oscillations and cerebral artery hemodynamics in humans. Microvasc Res 2011; 82(3): 304–310.

Tong

Hocke

Nickerson

, et al. Evaluating the effects of systemic low frequency oscillations measured in the periphery on the independent component analysis results of resting state networks. Neuroimage 2013; 76(1): 202–215.

10.

Tong

Hocke

Licata

, et al. Low-frequency oscillations measured in the periphery with near-infrared spectroscopy are strongly correlated with blood oxygen level-dependent functional magnetic resonance imaging signals. J Biomed Opt 2012; 17(10): 106004.

11.

Sassaroli

Pierro

Bergethon

, et al. Low-frequency spontaneous oscillations of cerebral hemodynamics investigated with near-infrared spectroscopy: a review. IEEE J Sel Top Quant Electron 2012; 18(4): 1478–1492.

12.

, et al. Asymmetry of peripheral vascular biomarkers in ischemic stroke patients, assessed using NIRS. J Biomed Opt 2020; 25(6): 1–16.

13.

, et al. A low-cost multichannel NIRS oximeter for monitoring. Systemic Low-Frequency Oscillations Neural Comput Appl 2020; 32(10): 15629–15641.

14.

Delpy

Cope

Zee

, et al. Estimation of optical pathlength through tissue from direct time of flight measurement. Phys Med Biol 1988; 33(12): 1433–1442.

15.

Kocsis

Herman

Eke

, et al. The modified Beer-Lambert law revisited. Phys Med Biol 2006; 51(5): 91–98.

16.

Aoki

Sato

Katura

, et al. Relationship of negative mood with prefrontal cortex activity during working memory tasks: an optical topography study. Neurosci Res 2011; 70(2): 189–196.

17.

Rutlen

Wackers

FJT

Zaret

, et al. Radionuclide assessment of peripheral intravascular capacity: a technique to measure intravascular volume changes in the capacitance circulation in man. Circulation 1981; 64(1): 146–152.

18.

Buhre

Weyland

Buhre

, et al. Effects of the sitting position on the distribution of blood volume in patients undergoing neurosurgical procedures. Br J Anaesth 2000; 84: 354–357.

19.

Jabot

. Passive leg raising for predicting fluid responsiveness: importance of the postural change. Intensive Care Med 2009; 35: 85–90.

20.

Dyer

, et al. Cubic-spline interpolation. 1. IEEE Instrum Meas Mag 2001; 4(1): 44–46.

21.

Chiarelli

Maclin

Fabiani

, et al. A kurtosis-based wavelet algorithm for motion artifact correction of fNIRS data. Neuroimage 2015; 112: 128–137.

22.

Bai

Xia

Wang

, et al. Fronto-parietal coherence response to tDCS modulation in patients with disorders of consciousness. Int J Neurosci 2018; 128(7): 587–594.

23.

Taga

Watanabe

Homae

. Spatial variation in the hemoglobin phase of oxygenation and deoxygenation in the developing cortex of infants. Neurophotonics 2018; 5(1): 011017.

24.