Determination of optimal catheter insertion depth in totally implantable venous access ports via three-dimensional reconstruction of cervicothoracic veins: a cross-sectional retrospective analysis.

Introduction
Totally implantable venous access ports (TIVAPs) are increasingly utilized in cancer management for chemotherapy administration and parenteral nutrition, due to their clinical advantages of enhanced portability, long indwelling duration, and reduced complication rates [1]. Precise positioning of the catheter tip during initial TIVAP implantation is critically important, as the device cannot easily be re-positioned after implantation, and demonstrated associations between the tip location and risk of complications [2]. Current guidelines recommend optimal tip placement at the cavoatrial junction (CAJ) or within the lower third of the superior vena cava (SVC) [3, 4]. Ectopic catheter tips can induce catheter dysfunction, venous thrombosis, arrhythmias, and other complications, adversely affecting patient outcomes [5, 6].
Intraoperative X-ray fluoroscopy is considered the gold standard for catheter tip localization [7]. However, it requires specialized equipment and involves ionizing radiation exposure, the magnitude of which depends on fluoroscopy duration and may accumulate for both patients and operators. Guidance using intracardiac electrocardiography (IC-ECG), a radiation-free alternative to fluoroscopy, enables real-time tip positioning, reduces the risk of malposition, and is cost-effective [8, 9]. Its contraindications include arrhythmias, P-wave abnormalities, or structural cardiac anomalies. In addition, while height formulas, surface landmarks, and imaging morphometry have been proposed to enhance the accuracy of catheter positioning, the accuracy of these tends to be poor [10, 11].
In recent years, three-dimensional (3D) vascular reconstruction technology has emerged as a pivotal tool, integrating tomography images to generate high-fidelity volumetric models. This technique provides clinicians with both spatially intuitive vascular mapping and quantitative precision, leading to its widespread adoption in vascular disease diagnosis, surgical navigation, and hemodynamic simulations [12–14]. While its application may depend on reconstruction time and reporting timelines, the integration of AI auto-segmentation algorithms further enhances the efficiency and accessibility of the 3D reconstruction processes [15]. However, to date, there have been no investigations of the use of 3D vascular reconstruction for predicting catheter length in TIVAP placement.
In this study, we propose a novel method for predicting catheter depth at the CAJ through 3D reconstruction of the cervicothoracic veins using non-contrast chest computed tomography (CT) scans. This method was verified by chest radiography and compared with the IC-ECG technique in a self-controlled design, assessing its feasibility and accuracy for the prediction of optimal catheter depth.

Materials and methods

Study design and datasets
This was a single-center, retrospective, cross-sectional study with a matched-pairs design. Consecutive female patients with breast cancer who underwent TIVAP placement for adjuvant chemotherapy were identified from the medical records of Affiliated Jinhua Hospital, Zhejiang University School of Medicine between January 2023 and December 2023. To enhance reporting transparency and methodological rigor, this study was designed and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for observational research. The matched-pairs design and predefined inclusion criteria were used to reduce selection bias and ensure comparability between predicted and optimal catheter depths across methods. Potential sources of bias inherent to retrospective analyses were mitigated by consecutive patient inclusion, standardized imaging acquisition protocols, and blinding of the radiologist performing 3D reconstruction to procedural and outcome data. The study concentration and outcome definitions were informed by previously published research on IC-ECG guidance, fluoroscopic validation, and radiographic localization of the cavoatrial junction, ensuring methodological consistency with existing literature. The inclusion criteria were as follows: (1) Age of 18–80 years; (2) Diagnosis of breast cancer and requiring TIVAP placement for adjuvant chemotherapy; (3) Internal jugular vein (IJV) access; (4) IC-ECG-guided insertion; (5) Availability of preoperative non-contrast chest CT and postoperative chest X-ray; and (6) Performed by a designated surgical team. Exclusion criteria included: (1) History of prior TIVAP placement or central venous catheterization; (2) Severe cardiac or pulmonary conditions that might interfere with IC-ECG; (3) Patients with advanced-stage metastatic cancer or those receiving palliative treatment. The final cohort was comprised of 50 patients who met these inclusion criteria.
The expected duration of TIVAP use in this cohort was anticipated to be between 6 and 12 months, corresponding to the duration of adjuvant chemotherapy. However, some patients may require continued use beyond this period, depending on the clinical course of their disease and treatment response. Device longevity and complication risks were particularly relevant to this cohort as TIVAPs are often intended for prolonged use.
Patient characteristics, including sex, age, height, weight, body mass index (BMI), actual catheter insertion depth, site of venous access, topographic distances, and TIVAP-associated complications, were retrospectively extracted from the hospital’s electronic medical record (EMR) system (HAITAI v4.0.207, Nanjing, China). Preoperative non-contrast chest CT in Digital Imaging and Communications in Medicine (DICOM) format and postoperative chest X-ray images were retrieved from the institutional picture archiving and communication system (PACS) (eWorldView v4.9, Ningbo, China).
The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Affiliated Jinhua Hospital, Zhejiang University School of Medicine (Approval No. [Res]2024-Ethics Review-16; Approval Date: February 2, 2024). The requirement for informed consent was waived due to the retrospective nature of the study and the use of fully anonymized historical data in which individual subjects could not be identified. Written informed consent for the publication of clinical images was obtained before image acquisition.

Surgical procedures
The Bard PowerPort™ Implantable Port System and its integrated components (Bard Access Systems Inc., Salt Lake City, UT, USA) were used for all patients. All procedures were performed by the same surgical team (Xiaotao Zhu, Lin Lv, Jialu Song, and Xiaohui Wang), in accordance with established guidelines [4, 16]. Preoperative assessment of the jugular venous anatomy and electrocardiography was performed, and written informed consent was obtained.
The key procedures were as follows: (1) Patients were positioned lying flat with the head turned to the contralateral side. The cervical puncture site and location of the infraclavicular port pocket were delineated by cutaneous markings. The topographic distances of the surface landmarks were measured. Standard ECG electrodes were applied to obtain the baseline cardiac rhythm. (2) Surgical field preparation and anesthesia with lidocaine infiltration were performed following routine practices. (3) Using ultrasound guidance in real-time, the left or right IJV was punctured at the level of the inferior thyroid border via the posterior approach. This approach offers several advantages, including a lower risk of carotid artery puncture, a more direct needle trajectory to the IJV, and reduced interference from the sternocleidomastoid muscle, thereby improving procedural safety and success rates. (4) A guidewire was then inserted and advanced, followed by placement of the introducer sheath. The central venous catheter was initially positioned at a depth of 10 cm. (5) A sterile RA ECG adapter was connected to the catheter hub, with documentation of the baseline P-wave amplitude during normal respiration. The catheter was advanced slowly until maximal P-wave amplitude was observed, at which point the optimal depth was established by retraction of the catheter by 1 cm. (6) A subcutaneous pocket was made in the chest wall. After tunneling from the chest incision to the puncture site, the catheter was trimmed to the predetermined length and secured to the port reservoir. (7) The final position of the catheter tip was confirmed on postoperative chest X-ray imaging, with the carina and vertebral bodies serving as landmarks to verify the tip location relative to the CAJ. All procedures were performed following internationally accepted guidelines for TIVAP implantation, and operator-related variability was minimized by restricting catheter placement to a single experienced surgical team, thereby strengthening internal validity.

Prediction methods
Two methods were employed to predict the optimal depth of catheter insertion, as follows:3D Reconstruction: The topographic distance (L1) from the puncture site to the ultrasound-guided junction of the left or right IJV and superior clavicular border was measured during surgery (Fig. 1a). Vascular segmentation was performed using 3D Slicer (v5.6.1) on the de-identified chest CT DICOM datasets (5-mm slice thickness) [17]. The segmentation was performed while blinded to the clinical data and was verified by a radiologist. The left or right IJV, superior vena cava, and proximal right atrium were segmented using the Segment Editor module employing the Paint and Fill Between Slices tools for manual refinement. Concurrently, the left or right clavicle and trunk of the right inferior pulmonary vein (RIPV) were segmented. 3D surface models were reconstructed for all structures, and the center line running from the superior clavicular border to the inferior RIPV margin was extracted using the Vascular Modeling Toolkit (SlicerVMTK) extension, with automatic quantification of the geodesic length (L2) (Fig. 1b). The predicted catheter depth was calculated as L1 + L2.

IC-ECG: All catheter placements were completed under IC-ECG guidance without the need for secondary surgical adjustments. The predicted catheter depth matched the actual catheter depth achieved during the placement.

The selection of both IC-ECG and CT-based 3D reconstruction as comparator methods was based on their established clinical use and prior validation in catheter tip localization studies, allowing meaningful comparison within a clinically relevant framework.

Observational parameters
The optimal catheter depth was defined at the CAJ. Catheter tip position was quantified as 2.4 vertebral body units (VBUs) below the carina on supine chest X-ray images, where one VBU was individually measured and defined as the vertical distance between the inferior endplates of adjacent vertebrae at the vertebral level of the carina. By using VBUs as an internal anatomical reference, this approach normalizes for inter-individual variations in vertebral size, allowing proportional catheter length adjustment according to patient anatomy (Fig. 1c) [16, 18]. The optimal catheter depth was calculated by adjusting the actual catheter depth using the distance between the catheter tip and the CAJ.
The signed distance (ΔL) between the predicted and optimal catheter depths was calculated for each method as follows: Predicted catheter depth - Optimal catheter depth (cm). The absolute distance (|ΔL|) was used to evaluate the predictive capability of catheter length estimation, eliminating directional bias. For enhanced data visualization, the signed distance was categorized into three strata: Shallow (ΔL < -1 cm), Optimal (-1 cm ≤ ΔL ≤ + 1 cm), and Deep (ΔL > + 1 cm).
Complications were classified according to their timing relative to the TIVAP placement procedure. Intraoperative complications were defined as adverse events occurring during the procedure, such as pneumothorax, accidental arterial puncture, arrhythmias, or air embolism. Short-term complications were defined as those occurring within 30 days post-procedure, such as bleeding, infection, wound dehiscence, or port rotation. Long-term complications were those occurring beyond 30 days, such as catheter occlusion, venous thrombosis, catheter dislodgement, fracture, or embolization. Procedural success was defined as the successful implantation of the TIVAP system with the catheter tip confirmed within a clinically acceptable range on postoperative X-ray and functional patency established intraoperatively, allowing for immediate clinical use. Additionally, functional success was assessed by the absence of TIVAP-related complications and the completion of chemotherapy cycles without the need for catheter repositioning or replacement.

Statistical analysis
Statistical analyses were performed using R software (v4.4.2; R Core Team, 2024). Data are expressed as means ± standard deviations (SD), medians [ranges], or numbers (percentages). Distribution normality was confirmed using the Shapiro-Wilk test. The ΔL values between two methods were compared using paired t-tests, while |ΔL| values were compared using Wilcoxon signed-rank tests. Distribution patterns were visualized using violin plots generated with the ggplot2 package. Inter-method agreement was assessed through Bland-Altman analysis using the Blandr package. The distributions of categorical variables were displayed through Sankey diagrams generated with the networkD3 package. Pearson correlation matrices were generated using the GGally package. The sample size was determined using the pwr package. Statistical significance was defined as p < 0.05.

Results
The study cohort comprised 50 patients who underwent TIVAP placement, of which 42 (84%) were accessed via the right IJV and 8 (16%) via the left IJV. All patients had a diagnosis of primary breast cancer and were undergoing adjuvant chemotherapy, with curative intent. Patients were selected based on the treatment regimens prescribed by the oncology team. The chemotherapy regimens were typically designed to achieve long-term remission and included combinations of agents, such as taxanes and anthracyclines. None of the patients in this study were undergoing chemotherapy with palliative intent, which might have altered catheter use duration and associated risks. The procedural success rate was 100% (50/50), as all devices were successfully implanted with catheter tips located within the clinically acceptable range and confirmed to be patent. During the six-month follow-up, all patients completed 4–8 chemotherapy cycles with no TIVAP-associated complications, such as catheter occlusion, venous thrombosis, and arrhythmias. Details of the patient characteristics are shown in Table 1.

The 3D reconstruction process was successfully completed in all patients. The observation parameters are summarized in Table 2. The mean optimal catheter depth was 15.1 ± 1.63 cm. The mean absolute deviations (MAD) of ΔL were 0.78 cm for 3D reconstruction and 0.92 cm for IC-ECG. The ΔL values did not differ significantly between the 3D reconstruction method and IC-ECG (-0.49 ± 0.95 cm vs. -0.82 ± 1.18 cm, p = 0.056, Fig. 2a), whereas the |ΔL| was significantly shorter for 3D reconstruction (0.70 [0, 2.60] cm vs. 1.00 [0.10, 3.10] cm, p = 0.007, Fig. 2b).

The results of the Bland-Altman analysis revealed no systematic bias (mean bias: 0.33 cm; 95% CI: -0.01 to 0.67), with limits of agreement (LoA) of -2.00 to 2.66 units. Three outliers exceeded the LoA without discernible patterns (Fig. 3). After stratification of the ΔL (shallow/optimal/deep), Sankey diagrams demonstrated bidirectional reclassification (Fig. 4). 3D reconstruction was found to predominate in optimal positions (70.0%), while IC-ECG showed the highest shallow-stratum representation (48.0%). Both methods had minimal deep-stratum proportions (3D reconstruction: 2.0%; IC-ECG: 4.0%).

Grouped correlation matrices were used to demonstrate pairwise relationships between ΔL and patient parameters (age, height, weight, BMI), stratified by the venous access site (right vs. left, Fig. 5). A weak positive correlation was observed in ΔL values from different methods both overall (Pearson’s r = 0.39, p < 0.01) and within the right-access subgroup (r = 0.42, p < 0.01). The IC-ECG ΔL showed a weak negative correlation with age overall (r = -0.30, p < 0.05), while the 3D reconstruction ΔL showed no significant correlations with age, height, weight, or BMI.

Discussion
This is the first investigation of the prediction of optimal catheter depth for TIVAP placement via 3D reconstruction of the cervicothoracic veins using non-contrast chest CT scans. The 3D reconstruction method demonstrated close agreement with the IC-ECG technique with no systematic bias (ΔL: p = 0.056; Bland-Altman mean bias: 0.33 cm). Notably, the 3D reconstruction method was more accurate in targeting the CAJ with reduced variability (both smaller SD and MAD) and smaller absolute values of the distances between the predicted and optimal catheter depths (|ΔL|: p = 0.007). Although 3D reconstruction demonstrated smaller absolute errors and reduced variability compared with IC-ECG, these differences should not be interpreted as definitive evidence of clinical superiority. The improvement in accuracy may partly reflect the controlled study environment and uniform imaging quality rather than a universally reproducible advantage across different institutions. Furthermore, the lack of statistically significant differences in signed distance (ΔL) may suggest that both methods perform comparably in achieving acceptable catheter tip positioning around the CAJ in routine clinical practice. The demonstrated accuracy and reliability of 3D reconstruction for predicting optimal catheter depth demonstrate that this method could be considered a valuable preoperative planning tool. For institutions equipped with routine non-contrast-enhanced chest CT, incorporating 3D reconstruction into the preoperative workflow may reduce reliance on intraoperative guidance methods that require specialized equipment (e.g., fluoroscopy) or are contraindicated in certain patients (e.g., IC-ECG in those with baseline P-wave abnormalities). This could influence device-selection protocols by favoring techniques that allow precise, individualized catheter-length determination before the procedure, thereby potentially lowering the risk of malposition-related complications. Regarding efficiency, the 3D reconstruction process required approximately 5 min per patient, presenting a feasible timeframe for preoperative planning without disrupting clinical workflow.
The recommended optimal catheter depth for chemotherapy was at the CAJ or within the lower third of the SVC. Previous studies have shown that shallow catheter placement increases the risk of dislodgement of the catheter into adjacent vessels, predisposes patients to venous wall damage due to inadequate drug dilution, and is an independent risk factor for venous thrombosis [5, 19, 20]. Conversely, deep placement may induce cardiac injury, resulting in arrhythmias, tricuspid valve damage, or even lethal pericardial tamponade [6, 21, 22]. Due to their design for extended indwelling periods and difficulties in adjustment post-placement, TIVAPs require greater precision of positioning than temporary central venous catheters (CVCs) or peripherally inserted central catheters (PICCs) to mitigate related complications. In clinical practice, the accurate prediction of catheter depth constitutes a critical step in TIVAP placement.
In this study, both the 3D reconstruction and IC-ECG methods targeted the optimal catheter depth at the CAJ, enabling precise measurement and standardized comparison. The final catheter depth was confirmed by chest X-ray. However, as the CAJ cannot be visualized directly on chest radiographs, the carina serves as a reliable landmark for CAJ localization [23]. Mehlon et al. reported a mean carina-to-CAJ distance of 40.3 mm (SD: 13.6 mm) based on CT angiography [24]. Similarly, Song et al. described a distance of 54.3 mm (SD: 9.7 mm) on CT angiography and demonstrated that the CAJ can be reliably estimated at 2.4 VBUs below the carina on chest X-rays using the thoracic spine as an internal reference [18]. This method was recommended by the Shanghai Expert Consensus on TIVAP placement and was implemented in the present study [16]. The results demonstrated good insertion accuracy for both methods. The IC-ECG technique was associated with a smaller mean ΔL, although the difference was non-significant, and higher shallow-stratum proportions compared to 3D reconstruction. This discrepancy is likely attributable to the safety protocol used during IC-ECG guidance, specifically, withdrawal of the catheter by 1 cm after maximal P-wave amplitude to prevent right atrial cannulation, an approach supported by Li et al. [11]. The reproducibility of the 3D reconstruction method, which is independent of operator experience and patient anthropometrics, provides an opportunity to standardize TIVAP placement training. By providing a quantifiable, image-based target, trainees can learn to anticipate the optimal catheter depth before entering the operating room. Integrating 3D reconstruction into simulation-based training modules can accelerate skill acquisition and improve consistency across operators, ultimately enhancing patient safety.
Several techniques have been developed for CAJ localization, classified into intraoperative guidance and preoperative prediction. Among the intraoperative methods, X-ray fluoroscopy provides a high degree of accuracy through real-time visualization of the catheter position using contrast dye. However, the application of this method is limited by its requirements for specialized equipment and concerns regarding radiation exposure. Manabu et al. reported that transesophageal echocardiography (TEE) achieves a precision in guidewire length measurement comparable to that of fluoroscopy [25]. Nevertheless, the clinical utility of TEE is restricted by its procedural complexity, associated patient discomfort, and risks of esophagogastric trauma. Overcoming these limitations, Marano et al. introduced a novel non-invasive empirical-ultrasonographical approach for PICC placement, known as the “Marano index.” As a less invasive alternative to TEE, this technique utilizes transabdominal ultrasound to visualize the catheter tip via the subxiphoid window, achieving a remarkable 94% accuracy without fluoroscopy [26]. In contrast, IC-ECG monitors the position of the catheter tip via changes in the characteristic P-wave morphology detected on electrocardiographic monitors. Maximal P-wave amplitude indicates optimal positioning at the CAJ or the possibility of slight entry into the right atrium [27]. This technique achieved an accuracy of 95.4%, as reported by Pittiruti et al., and an accuracy of 99.3% by Li et al., without additional trauma or costs [28, 29]. However, it cannot be used in patients with baseline P-wave abnormalities, with 0.7% of cases exhibiting no anticipated P-wave amplitude increase despite normal baseline P-waves [29]. It is also important to recognize that both IC-ECG and ultrasound guided measurements are inherently operator dependent. The high procedural success and accuracy observed in this study may therefore reflect institutional experience and adherence to standardized techniques, potentially limiting generalizability to centers with different levels of expertise. The inability to perform preoperative planning when using the above approaches increases surgical uncertainty and extends the operating time. Methods that permit preoperative assessment, such as height-based or surface landmark-based formulas, have been used in the past [30–32]. However, although these are easy to use, their accuracy is poor due to susceptibility to variations in puncture sites and patient body habitus, sex, and ethnicity. Several studies have reported that chest CT-based morphometric measurements predict catheter depth more accurately than traditional formulas but less accurately than IC-ECG [11, 33]. For instance, Dong et al. utilized the vertical distance between the right sternoclavicular joint and the right cardiac apex border as a key metric in PICC placement [33].
While this study presented favorable outcomes with no TIVAP-associated complications during the six-month follow-up period, it is important to contextualize these findings with respect to established benchmarks. Previous studies have reported complication rates ranging from 2% to 12% for venous thrombosis, catheter occlusion, and arrhythmias in similar cohorts [5, 19, 20]. The absence of complications in our cohort likely reflects a multifactorial advantage. Standardization of the surgical technique is a critical determinant of outcomes. As highlighted by Abou-Mrad et al. in a recent 5-year monocentric analysis, strict adherence to standardized surgical protocols and patient management can achieve a procedural success rate of 100% while ensuring long-term device safety [34]. Our findings are consistent with these observations, suggesting that the absence of complications in our cohort, in which insertions were performed under IC-ECG guidance, is attributable primarily to rigorous protocol adherence and operator expertise. Although our 3D prediction method demonstrates high theoretical accuracy, the procedure’s fundamental safety remains strongly dependent on standardized practice and technical proficiency. However, these results should be interpreted with caution. The absence of complications may also reflect specific cohort characteristics (e.g., exclusion of palliative cases) or the limited sample size rather than an inherently superior outcome of the 3D method itself. The relatively small sample size and the six-month follow-up period may have limited the ability to detect infrequent or delayed complications, such as late-onset thrombosis or catheter dysfunction. Due to the absence of a contemporaneous comparator group, the reported safety outcomes cannot be directly benchmarked against other techniques (e.g., fluoroscopy-guided cohorts). Future studies incorporating such control groups are warranted to provide a more robust evaluation. Regarding follow-up, the high predictive accuracy of this method may allow for more focused surveillance. While post-procedural X-ray remains standard for safety, the assurance of optimal tip positioning reduces the likelihood of late complications such as thrombosis or dysfunction. This reliability could streamline long-term maintenance protocols, allowing clinicians to focus follow-up resources on patients with identified anatomical complexities rather than routine troubleshooting of malposition concerns.
The vascular 3D reconstruction model is a form of imaging morphometry. The incorporation of comprehensive 3D spatial data of the vasculature enhances the accuracy of this approach compared to conventional CT slice measurements and reconstructed planar vascular analyses [35]. As percutaneous access-to-vessel distances cannot be measured within 3D models, we established the jugular-clavicular junction as a landmark, segmenting the catheter pathway into extravascular and intravascular compartments. The location of this landmark was defined using intraoperative ultrasound for precise extravascular measurement. For intravascular CAJ localization, the inferior border of the right pulmonary artery, which is easily identified, was selected as the anatomical landmark. While contrast-enhanced chest CT facilitates vascular reconstruction, the present study confirms the clinical viability of non-contrast CT strategies. This eliminates the need for additional examinations since non-contrast chest CT scans are already integral to oncological evaluation. This protocol improves the feasibility and reproducibility of the 3D reconstruction method. Theoretically, this method is broadly applicable and is not sensitive to patient characteristics and venous access sites. Our data confirmed that ΔL was not associated with age, height, weight, or BMI. In addition, we observed a moderate positive correlation (r = 0.394) between the 3D reconstruction and IC-ECG ΔL values, particularly when both methods yielded negative values. This concordant underestimation suggests shared influencing factors, possibly including gravity-induced port migration, tissue compliance-mediated catheter redistribution, connection-related truncation errors, and the effects of diaphragmatic motion [36, 37]. These variables are also likely to contribute to outlier measurements and warrant consideration during catheter placement.
This study has several limitations. First, from a methodological standpoint, the retrospective and monocentric nature of the study restricts the generalizability of the results. As institutional practices, patient demographics, and healthcare infrastructure vary across centers, our findings which were derived from a single experienced surgical team may not be universally reproducible. Furthermore, the retrospective design limits control over potential confounding factors. Second, regarding patient characteristics, the cohort consisted entirely of female patients with breast cancer, predominantly utilizing right IJV access. Validation in male cohorts is necessary. Third, the applicability to other access sites is constrained by imaging protocols. While feasible for subclavian or axillary veins within the standard chest CT range, application to cephalic or basilic veins would require adapting the model for external distal measurement to avoid the additional radiation exposure associated with extending the CT scan field. Fourth, there are limitations inherent to the technique and surveillance. While accurate for depth prediction, this method cannot detect intraoperative malpositioning, necessitating combination with intraoperative guidance. Additionally, although no complications were observed, the six-month follow-up may be insufficient to capture late-onset events or subclinical complications managed outside our center, leading to potential under-ascertainment. Finally, operator dependency remains a consideration. The accuracy of both ultrasound-based landmark identification and IC-ECG interpretation can vary by expertise level. Consequently, larger multicenter prospective studies with longer follow-up are warranted to disentangle the specific contribution of this prediction method from confounding factors such as operator skill, and to confirm its robustness across diverse populations and clinical settings.

Conclusions
This study establishes non-contrast chest CT-based 3D vascular reconstruction as a clinically viable method for predicting the optimal catheter depth in TIVAP placement. Compared to IC-ECG guidance, the 3D reconstruction technique demonstrated comparable precision with reduced absolute distance (|ΔL| 0.70 cm vs. 1.00 cm, p = 0.007) and superior accuracy in targeting the CAJ. It is broadly applicable and eliminates the need for additional examinations, as routine chest CT scans are integral to oncological evaluations. Validation in cohorts with males and with different venous access sites is warranted to confirm the generalizability of the findings. When combined with intraoperative guidance, this technique may optimize surgical efficiency and reduce the risk of catheter malpositioning, though routine postoperative X-ray remains recommended for safety confirmation.