TCD is a noninvasive technique based on the Doppler effect. The underlying principle is that when an ultrasound wave emitted from the probe encounters moving blood, primarily red blood cells, the frequency of the reflected wave changes. The frequency increases when blood moves toward the probe and decreases when it moves away. TCD devices detect this frequency shift and, together with the angle of insonation, calculate blood flow velocity within the targeted vessel. In clinical settings, TCD provides a series of essential parameters that represent the cerebral hemodynamic state (Table 1) [8].

(1) Peak systolic velocity (PSV): The maximum flow velocity during systole, reflecting cardiac output and vascular compliance.

(2) End-diastolic velocity (EDV): The minimum flow velocity during diastole, primarily influenced by distal vascular resistance.

(3) Mean flow velocity (MFV): The average velocity throughout the cardiac cycle, proportional to cerebral blood flow (CBF) and commonly used to evaluate cerebral perfusion.

(4) Pulsatility index (PI): Calculated as PI = (PSV - EDV) / MFV; serves as an indirect indicator of distal vascular resistance and intracranial pressure (ICP).

(5) Resistive index (RI): Calculated as RI = (PSV - EDV) / PSV; similar to PI, it is used to assess vascular resistance.

These parameters collectively constitute a comprehensive hemodynamic monitoring framework. For instance, a pronounced decrease in MFV accompanied by a substantial increase in PI and RI following anesthetic induction in an elderly patient strongly suggests inadequate cerebral perfusion or elevated intracranial pressure, warranting prompt clinical intervention.

To obtain stable and reliable TCD signals, clinical procedures should follow standardized protocols [9]. The most commonly used site is the transtemporal window, located above the ear, which permits ultrasound transmission through the thinnest portion of the temporal bone. This window enables the assessment of blood flow in the middle cerebral artery (MCA), anterior cerebral artery, and posterior cerebral artery. Among these, the MCA serves as a primary indicator of global cerebral hemodynamics due to its consistent anatomical position and extensive perfusion territory. Other commonly used acoustic windows include the transoccipital window, which provides access to the vertebrobasilar system, and the transorbital window, which allows visualization of the ophthalmic artery and internal carotid siphon. In specific procedures such as carotid endarterectomy (CEA), blood flow changes detected through these windows are monitored to evaluate the operation’s effect on cerebral perfusion. Despite its advantages, TCD application continues to face certain limitations. Inadequate acoustic windows—often resulting from skull thickening—occur in approximately 10-20% of elderly Asian patients, thereby restricting its widespread use [10]. Furthermore, TCD is highly operator-dependent, necessitating specialized training to achieve optimal probe placement and signal acquisition [11].

TCD technology is progressing toward greater automation and integration with AI, aiming to reduce operator dependency and enhance clinical applicability. Robotic TCD: Robotic systems can autonomously identify intracranial acoustic windows and maintain stable probe positioning, thereby minimizing technical variability and improving the success rate of signal acquisition. Clinically, this advancement allows consistent and reliable monitoring even in patients with suboptimal acoustic windows [11]. AI-assisted analysis: AI algorithms can automatically detect abnormal TCD patterns—such as high-intensity transient signals (HITS)—and continuously track flow velocity trends in real time. This functionality supports clinicians by providing early alerts regarding potential cerebral perfusion abnormalities and assisting with decision-making during perioperative management [12,13]. These technological innovations indicate that TCD is evolving from a skill-dependent diagnostic modality into a standardized, automated, and clinically versatile instrument, thereby enhancing precision brain monitoring in perioperative settings.

CA is a vital physiological mechanism that maintains CBF within a stable range despite fluctuations in mean arterial pressure (MAP), typically between 60 and 150 mmHg. This regulation relies on the integrated actions of multiple mechanisms [14], including the myogenic response (smooth muscle contraction in vessel walls), neurogenic control (autonomic nervous system activity), and metabolic regulation (changes in brain metabolites such as carbon dioxide [CO₂] and nitric oxide [NO]). However, CA function declines significantly in the elderly due to several interrelated factors [15]:

(1) Vascular structural alterations: Atherosclerosis leads to arterial stiffening and reduced elasticity, thereby weakening vascular responsiveness to blood pressure variations.

(2) Impaired neurovascular coupling: The coordination between neuronal activity and regional CBF diminishes, limiting the brain’s ability to rapidly adjust local blood flow according to metabolic demand.

(3) Endothelial dysfunction: An imbalance between vasodilatory mediators (e.g., NO) and vasoconstrictive substances produced by endothelial cells disrupts normal vascular regulation.

These alterations collectively shift the autoregulatory curve in elderly patients toward a higher blood pressure range and narrow its effective operating bandwidth. Consequently, older adults require a higher MAP to sustain adequate cerebral perfusion and exhibit a reduced tolerance to blood pressure fluctuations. Conventional blood pressure management targets may therefore be insufficient to ensure optimal cerebral perfusion in this population, increasing the risk of cerebral hypoperfusion during the perioperative period [14].

The effects of anesthetic agents on cerebral hemodynamics are multifaceted and vary according to the pharmacological class of the drug. Clinically, TCD enables real-time monitoring of CBF velocity and autoregulatory dynamics [16].

(1) Inhalational anesthetics: Volatile agents such as sevoflurane and isoflurane primarily act by dose-dependently suppressing neuronal activity and reducing cerebral metabolic rate. However, these agents also exert potent vasodilatory effects, particularly on cerebral vessels [17]. In younger patients with intact CA, this vasodilation is typically counterbalanced by autoregulatory mechanisms. In elderly patients with impaired CA, however, this effect can lead to a perfusion-metabolism uncoupling, making CBF more dependent on systemic blood pressure and further compromising autoregulatory function.

(2) Intravenous anesthetics: Intravenous agents such as propofol and etomidate generally decrease CBF by lowering the cerebral metabolic rate. This metabolism-perfusion coupling may be beneficial in elderly patients with impaired CA, as it reduces the brain’s metabolic and perfusion demands [18]. Nevertheless, when accompanied by a marked reduction in systemic blood pressure, these agents may still precipitate cerebral hypoperfusion.

(3) Neuromuscular blockers and opioids: Most neuromuscular blocking agents exert minimal effects on CBF. Opioids (e.g., fentanyl) slightly decrease CBF by reducing cerebral metabolism, although their impact on autoregulatory function remains negligible.

The use of real-time TCD monitoring during anesthetic induction, maintenance, and emergence allows clinicians to continuously evaluate how anesthetic agents influence a patient’s CA function [19]. Early alterations in MFV and PI on TCD can serve as early indicators of cerebral hypoperfusion, enabling timely interventions such as anesthetic titration or vasopressor administration to preserve cerebral perfusion and protect brain function.

Perioperative microembolism plays a pivotal role in the development of postoperative stroke and cognitive impairment. These microemboli—composed of air, fat, or platelet aggregates—enter the cerebral circulation and occlude small vessels, resulting in localized microinfarctions. Real-time detection of HITS—brief, high-amplitude spikes on the TCD waveform whose frequency and count reflect the microembolic load—enables clinicians to assess embolic risk and implement timely preventive interventions during surgery [26]. Continuous HITS monitoring is particularly valuable during procedures such as CEA, cardiac valve replacement, and major orthopedic operations, including total hip arthroplasty [27]. Evidence indicates that a high intraoperative HITS burden is an independent predictor of POD and POCD [28]. These microemboli may obstruct cerebral microvessels and induce neuroinflammatory responses, ultimately contributing to postoperative cognitive decline. Accordingly, real-time HITS surveillance provides actionable feedback, allowing surgeons to minimize embolic release and assisting anesthesiologists in evaluating the efficacy of thromboprophylactic strategies.

Beyond microembolic events, both insufficient and excessive CBF can be detrimental [20].As previously described, impaired CA in older adults increases their susceptibility to hypotension. Early reductions in MFV serve as sensitive indicators of cerebral hypoperfusion, allowing clinicians to initiate corrective measures before systemic hypotension develops. Sustained or recurrent cerebral hypoperfusion during surgery represents a major risk factor for POD and POCD [29], as even brief episodes of reduced perfusion may lead to irreversible neuronal injury [30]. Conversely, cerebral hyperperfusion can disrupt the blood-brain barrier and precipitate cerebral edema, particularly following CEA. Postoperative monitoring of MCA flow velocities enables early recognition of abnormally elevated values, indicating a heightened risk of hyperperfusion and guiding precise blood pressure management to prevent postoperative stroke or hyperperfusion syndrome.

In addition to isolated hypo- or hyperperfusion states, emerging evidence suggests that fluctuations in CBF during the perioperative period may themselves constitute an independent risk factor for POCD [20]. TCD enables continuous monitoring of subtle flow variations, including high-frequency oscillations in MFV even within normal MAP ranges. Such variability is increasingly recognized as a marker of impaired or unstable CA, potentially associated with microvascular injury, neuronal dysfunction, and blood-brain barrier compromise [21,31]. By quantifying CBF variability in real time, anesthesiologists can identify patients at increased risk of cognitive decline, optimize hemodynamic and perfusion strategies, and maintain greater stability in CBF. This proactive, data-driven management approach may reduce perioperative neurological complications and support brain-protective anesthesia practices.

The principal advantage of TCD lies in its capacity to dynamically assess a patient’s CA function rather than presuming its normality. Assessment typically depends on evaluating the relationship between MFV and MAP [36]. Under ideal conditions, when MAP fluctuates within the autoregulatory range, MFV remains relatively stable, and the correlation between the two parameters is weak. However, once MAP falls outside the autoregulatory limits, MFV becomes positively correlated with MAP—meaning that a decrease in blood pressure leads to a proportional decline in CBF [31]. Building upon this principle, researchers have developed several methodologies to quantify CA function and determine an individual’s optimal MAP [37]:

(1) Cerebral autoregulation index (ARI): ARI is among the most widely utilized quantitative indices for assessing dynamic CA (dCA). It is a dCA monitoring method based on a second-order differential model [38]. ARI values range from 0 to 9, where 0 represents a complete loss of dCA (CBF passively follows MAP changes), and 9 indicates nearly intact dCA (CBF rapidly returns to baseline following a step change in MAP). ARI thus provides an intuitive and clinically applicable approach for monitoring dCA, remaining one of the most established tools in both research and perioperative practice.

(2) Correlation analysis: The mean flow index (Mx) is a time-domain metric used to assess dCA [39]. It quantifies the correlation between spontaneous fluctuations in arterial blood pressure (ABP)—or cerebral perfusion pressure—and CBF velocity in the MCA (MCAv), as measured by TCD. A higher Mx value denotes a stronger positive correlation, reflecting impaired autoregulatory function, whereas a lower or negative Mx value indicates preserved autoregulation. Typically, continuous recordings of ABP and MCAv are averaged over sequential time intervals, or “blocks,” each lasting approximately 10 s; however, some studies have adopted shorter durations, such as 5 s [39]. Roughly 30 consecutive blocks constitute an “epoch,” during which the Pearson correlation coefficient between MAP and MCAv is calculated, producing a single Mx value that can be updated continuously. This analytical approach enables near real-time evaluation of dCA and is widely applied in both research and clinical contexts to optimize cerebral perfusion management.

After determining the optimal MAP, TCD also serves as a real-time feedback tool for guiding clinical interventions. When a patient’s MAP falls below the identified optimal level and TCD demonstrates a corresponding reduction in MFV, it signals the need for vasopressor administration.

(1) Drug selection: TCD assists in evaluating the differential effects of various vasopressors on CBF [40]. For instance, phenylephrine primarily increases blood pressure through peripheral vasoconstriction with minimal direct influence on cerebral vessels, whereas ephedrine may enhance CBF indirectly by augmenting cardiac output. By continuously tracking MFV changes, clinicians can determine which vasopressor more effectively improves cerebral perfusion in a given patient.

(2) Dose adjustment: During vasopressor titration, TCD provides real-time feedback on MFV dynamics, enabling anesthesiologists to identify the optimal dosing range. Stabilization of MFV values accompanied by an increase in the ARI indicates that the current hemodynamic management strategy is effective, thereby avoiding unnecessary vasopressor escalation. This precision-guided approach minimizes cardiac workload and lowers the risk of hypertensive or hemorrhagic complications.

In summary, TCD transforms blood pressure management in elderly patients from an empirical, experience-based practice into a precise, data-driven, and individualized process. This approach not only prevents cerebral hypoperfusion but also mitigates the risks associated with overtreatment, forming a critical foundation for precision anesthesia and personalized brain protection.

In addition to its intraoperative applications, TCD can serve as a noninvasive and practical tool for preoperative risk assessment. Evaluation of CVR capacity can be performed using the CO₂ reactivity test. CVR refers to the ability of cerebral vessels to constrict or dilate in response to stimuli such as changes in cerebral perfusion pressure or arterial partial pressure of CO₂ (PaCO₂) [41]. The specific ability of cerebral vessels to respond to variations in arterial CO₂ levels is termed cerebrovascular CO₂ reactivity (CVR-CO₂). Assessment of CVR-CO₂ provides valuable information about cerebrovascular regulation and collateral circulation, which is crucial for the early diagnosis, treatment, and prognosis of several neurological conditions, including stroke and cognitive impairment. Clinically, various imaging modalities are used in combination with physiological tests such as the breath-holding test, CO₂ inhalation, or acetazolamide challenge to quantify CVR-CO₂. In 1982, Norwegian researcher Aaslid et al. [7] first applied TCD in clinical practice. Its noninvasive, cost-effective, reliable, and user-friendly characteristics have since facilitated its widespread use in screening for cerebrovascular disorders. CVR-CO₂ is known to be altered under certain pathophysiological conditions. Multiple studies have demonstrated that patients with Alzheimer’s disease and vascular dementia exhibit significantly lower CVR-CO₂ compared with healthy controls. Following approximately five weeks of treatment with anti-dementia medications, CVR-CO₂ values showed marked improvement [42,43]. Similarly, Lee et al. [44] reported that in patients with reduced cerebral perfusion pressure, cerebrovascular reactivity to CO₂ and vasodilators was diminished, preventing CBF from reaching normal levels. These studies collectively demonstrate that CVR-CO₂ can assist clinicians in comprehensively evaluating cerebrovascular regulatory capacity and determining the degree of risk associated with reduced cerebral perfusion pressure in individual patients. Integrating TCD monitoring data with conventional risk factors—such as patient age, comorbidities, and surgical type—shows great potential for the development of a more comprehensive preoperative neurological risk prediction model. Such an integrative approach may enable the early identification and management of high-risk patients through anesthetic optimization, enhanced intraoperative monitoring, and the implementation of targeted neuroprotective strategies.

NIRS indirectly evaluates the balance between cerebral oxygen supply and demand by measuring hemoglobin oxygenation levels, thereby providing regional cerebral oxygen saturation (rSO₂) data [47]. Combining NIRS with TCD enables a more comprehensive assessment of cerebral oxygen supply-demand dynamics [48,49].

(1) Identifying supply-demand mismatch: When a reduction in MFV detected by TCD coincides with a concurrent decline in rSO₂ measured by NIRS, this strongly suggests severe cerebral hypoperfusion requiring immediate intervention, such as vasopressor administration or blood transfusion.

(2) Differential diagnosis: If TCD reveals a decrease in MFV while NIRS-derived rSO₂ remains stable, this pattern may indicate compensatory cerebral adaptation through an increased oxygen extraction ratio. Although the immediate risk is lower, vigilant monitoring remains necessary. Conversely, if MFV measured by TCD increases but rSO₂ on NIRS fails to rise proportionally, it may indicate microvascular dysfunction or impaired oxygen diffusion.

This integrated multimodal approach allows clinicians to distinguish whether inadequate cerebral perfusion results primarily from decreased blood flow, reduced oxygen-carrying capacity (e.g., anemia), or microcirculatory impairment, thereby guiding targeted interventions. Compared with other perioperative brain monitoring techniques, each modality provides distinct yet complementary data: BIS and EEG primarily reflect cortical electrical activity, NIRS measures regional oxygenation, and TCD uniquely quantifies dynamic CBF while detecting microembolic signals via HITS monitoring. Integrating TCD with multimodal platforms that include NIRS and BIS enhances clinical decision-making by combining hemodynamic, metabolic, and electrophysiological perspectives. This comprehensive strategy advances precision in cerebral monitoring and supports personalized brain-protection protocols in geriatric anesthesia.

In the future, the integration of TCD is expected to extend beyond traditional neuromonitoring modalities. For example, it can be incorporated into research on cerebrospinal fluid (CSF) and serum biomarkers. Perioperative brain injury is often associated with elevated levels of neuronal injury biomarkers [50], such as neuron-specific enolase and S100B protein, in the serum or CSF. Correlating intraoperative TCD-derived parameters—including microembolic load and hypoperfusion duration—with postoperative biomarker concentrations may help validate the predictive utility of TCD for perioperative brain injury, thereby establishing the foundation for more comprehensive risk assessment models. Ultimately, integrating these technologies into a multimodal monitoring platform enhanced by AI-assisted analysis is likely to define the next phase of precision anesthesia. Such a platform would display real-time data from TCD, NIRS, and EEG, automatically calculate indices such as the ARI, detect abnormal hemodynamic or metabolic patterns, and generate alerts. The anesthesiologist could then interpret these alerts and synthesize multimodal data to perform a comprehensive evaluation—for instance, determining whether a complication arises from impaired blood flow (TCD decrease) or oxygenation deficit (NIRS decrease). This integrative, AI-assisted approach has the potential to transform anesthetic management from a reactive to a proactive model, ensuring optimal cerebral protection in elderly surgical patients.

Recent advances in robotic probe positioning and AI-based waveform analysis show considerable promise in reducing operator dependence and enhancing monitoring stability. These technologies can automatically identify optimal acoustic windows, maintain probe alignment, and generate real-time alerts for abnormal cerebral flow patterns. However, their clinical feasibility and safety remain under investigation. Current research primarily consists of feasibility or proof-of-concept studies, with large-scale validation across diverse perioperative populations still lacking. Additionally, issues related to probe fixation safety, algorithm transparency, and integration into existing perioperative monitoring workflows must be addressed before routine clinical implementation.

The development of compact, portable, or wearable TCD systems may enable continuous, long-term hemodynamic monitoring in post-anesthesia care units, intensive care units, or even in outpatient and home-care settings. Such advancements would extend the role of TCD beyond the operating room, expanding opportunities for continuous perioperative and post-discharge brain protection in elderly patients.

To fully realize the potential of TCD, future research should prioritize the following directions:

(1) Conducting large, multicenter randomized controlled trials to validate the relationship between TCD-derived parameters and perioperative neurological outcomes.

(2) Assessing robotic and AI-assisted TCD systems across diverse patient populations, with emphasis on safety, reproducibility, and seamless integration into perioperative workflows.

(3) Developing standardized multimodal monitoring protocols that combine TCD with EEG, BIS, NIRS, and biomarker analysis to achieve personalized brain-protection strategies.

(4) Investigating long-term cognitive outcomes in relation to perioperative cerebral perfusion variability and cumulative microembolic load.

In summary, TCD is poised to serve as an indispensable “eye” in geriatric anesthesia, providing critical support for achieving precision anesthesia and individualized brain protection.