After branching from the external iliac artery, the DIEA travels superiorly within the rectus sheath. In a retrospective CTA study of 184 patients, the DIEA most commonly originated medially, descended briefly, and then turned cephalad. The average descending segment measured 11.29 mm^[9]. Prior abdominal surgery was associated with a longer descending course (12.22 mm vs. 9.86 mm, P < 0.01), which may complicate pedicle dissection^[9]. The vessel’s branching pattern - Type I (single trunk), Type II (bifurcating), or Type III (trifurcating) - also has practical implications. Trifurcating patterns, for example, are linked to greater transverse spread (P = 0.0002) in a cadaveric study and may require more extensive rectus muscle dissection^[10]. These anatomical observations, however, do not establish differences in clinical morbidity.

Perforators are typically grouped into medial and lateral rows, each with distinct perfusion characteristics^[11]. Medial row perforators tend to provide more central perfusion and more reliably supply zone IV, whereas lateral row perforators primarily perfuse ipsilateral tissue and less frequently cross the midline^[12,13]. Correspondingly, medial row perforators support a larger vascular territory on average (296 cm² vs. 196 cm²)^[13]. From a technical standpoint, lateral row perforators are often easier to dissect, as they follow a more direct intramuscular path in the majority of cases (79.2% vs. 18.2% for medial row, P = 0.001)^[14]. However, this comes with trade-offs. In a retrospective analysis of 409 DIEP flaps, lateral-row harvest was associated with lower odds of fat necrosis [odds ratio (OR) of 0.29, P = 0.001], whereas lateral-row or combined medial-and-lateral-row harvest was associated with higher odds of bulge or hernia than medial-row harvest (OR 3.21; P = 0.05)^[15].

Perforators are typically grouped into medial and lateral rows, each with distinct perfusion characteristics^[11]. Medial row perforators tend to provide more central perfusion and more reliably supply zone IV, whereas lateral row perforators primarily perfuse ipsilateral tissue and less frequently cross the midline^[12,13]. Correspondingly, medial row perforators support a larger vascular territory on average (296 cm² vs. 196 cm²)^[13]. From a technical standpoint, lateral row perforators are often easier to dissect, as they follow a more direct intramuscular path in the majority of cases (79.2% vs. 18.2% for medial row, P = 0.001)^[14]. However, this comes with trade-offs. In a retrospective analysis of 409 DIEP flaps, lateral-row harvest was associated with lower odds of fat necrosis [odds ratio (OR) of 0.29, P = 0.001], whereas lateral-row or combined medial-and-lateral-row harvest was associated with higher odds of bulge or hernia than medial-row harvest (OR 3.21; P = 0.05)^[15].

DIEP flap perforators commonly cluster in the periumbilical region^[11,16]. In a CTA study of 180 patients, each hemiabdomen had an average of 2.9 ± 1.6 perforators, with the majority (80.5%) located below the umbilicus, and larger-caliber vessels were closer to the umbilicus^[16]. This distribution makes the periumbilical area a key focus during perforator selection.

Most DIEP perforators are musculocutaneous (33%-100%), and in many cases, they have a relatively short intramuscular course (< 4 cm) (61%-80%), though longer courses are not uncommon (9%-26%)^[8]. Septocutaneous variants, such as paramedian or tendinous perforators, are generally preferred when present, given more straightforward dissection^[8]. The length and complexity of the intramuscular course are closely linked to operative time. Longer or stepwise paths can significantly prolong dissection, whereas paramedian septocutaneous perforators often allow more efficient elevation^[17,18].

Venous drainage occurs through both the deep venous system (venae comitantes) and the superficial inferior epigastric vein (SIEV), though the latter has historically been underemphasized^[12,19]. In a retrospective MRA study of 59 patients, communication between the superficial and deep venous systems was identified in 117/118 hemiabdomens; mean diameters of the SIEV and dominant venous perforator were 3.8 and 2.8 mm, respectively^[20].

Incorporating both venous systems can potentially improve outcomes. A retrospective series comparing 130 flaps with deep-system-only drainage and 114 with combined superficial and deep drainage showed lower flap loss in the dual-system group (2.5% in deep-system-only vs. 0% in dual-system; P = 0.031)^[21]. A meta-analysis of 13 comparative studies demonstrated that venous supercharging with the superficial system reduces venous congestion [risk ratio (RR) of 0.12, 95% confidence interval (CI): 0.04-0.34], partial flap necrosis (RR 0.50, 95% CI: 0.30-0.84), total flap necrosis (RR 0.31, 95% CI: 0.11-0.85), and return to surgery for perfusion-related complications (RR 0.45, 95% CI: 0.21-0.99)^[22]. Although these findings suggest a benefit, a subsequent systematic review rated the evidence of reducing flap complications with prophylactic venous supercharging as low certainty without clinical signs of venous congestion^[23]. Evidence supporting preoperative imaging or measurements to identify patients who would benefit from augmentation was of very low certainty^[23]. Future studies may be needed to clarify the evidence. Nevertheless, preoperative imaging can help identify patients who may benefit from planned venous augmentation^[22].

These anatomical factors, such as perforator location, intramuscular course, proximity to the umbilicus, and venous anatomy, directly impact pedicle length, ease of dissection, and flap perfusion reliability. Thoughtful perforator selection can improve efficiency while minimizing donor-site morbidity^[15-17]. In this context, CTA-based planning has demonstrated high accuracy for perforator localization, with reported sensitivity approaching 99% and mean localization error within 1 cm^[24].

Of note, most anatomical data summarized in this section are derived from cadaveric studies or retrospective imaging analyses. While these investigations have substantially improved understanding of perforator anatomy, relatively few studies directly correlate individual anatomic characteristics with clinically meaningful reconstructive outcomes.

CTA is the most commonly used imaging modality for DIEP flap planning, as it provides 3D visualization of vascular anatomy^[2,24]. A 2025 systematic review and meta-analysis found that CTA-guided reconstruction is associated with shorter operative times (by 65 min), lower flap loss (RR 0.37; 95% CI: 0.18-0.78), and fewer reoperations (RR 0.55; 95% CI: 0.33-0.89) compared to Doppler ultrasound^[1]. CTA also demonstrates high accuracy for perforator localization, with 99% sensitivity and a mean error ≤ 1 cm^[24].

CTA is particularly effective for assessing intramuscular perforator course (95% accuracy) and identifying superficial venous connections (97% accuracy), the absence of which has been associated with increased flap morbidity (P = 0.009)^[7]. It also clearly delineates DIEA branching patterns and is especially useful in patients with prior abdominal surgery, where anatomy may be altered^[2].

Limitations include exposure to ionizing radiation, the need for iodinated contrast, and a lack of real-time hemodynamic assessment^[6]. CTA is also less accurate for perforator caliber measurement compared to intraoperative findings (P < 0.0001), which may affect perforator selection^[7].

CDU has gained renewed interest due to the general increased adoption of ultrasound in a wide array of procedural and non-procedural practices, as well as concerns regarding radiation exposure^[6]. It enables real-time, dynamic assessment without radiation or contrast and allows for the evaluation of flow characteristics and pulsatility^[4]. CDU can also be performed by the operating surgeon given appropriate training and expertise^[4]. In a comparative study, CDU caliber measurements did not differ significantly from intraoperative measurements (P = 0.33), whereas CTA measurements showed a statistically significant difference (P < 0.0001)^[7]. A randomized trial demonstrated that CDU is non-inferior to CTA+CDU for perforator localization, with smaller deviation from intraoperative position (0.6 cm vs. 1.0 cm, P < 0.001)^[3]. Cohort data also show a stronger correlation with intraoperative perforator selection for CDU (r = 0.9987) compared to CTA (r = 0.01)^[6]. Limitations include operator dependence, lower accuracy in assessing intramuscular course (84% vs. 95% for CTA) and superficial venous connections (80% vs. 97%), and lack of a comprehensive anatomical overview^[7]. While observational studies tend to favor CTA, randomized data show no clear difference, suggesting that CTA’s perceived advantage may be influenced by confounding factors^[1].

MRA provides high-resolution, cross-sectional imaging without radiation. Meta-analyses show the strongest agreement with intraoperative findings (mean 97%; 95% CI: 86%-100%) among available modalities^[25]. It offers excellent soft tissue contrast and can evaluate both arterial and venous anatomy^[2]. Its drawbacks, however, include longer scan times, higher cost, limited availability, and contraindications in patients with certain implants. Although MRA demonstrates excellent diagnostic performance, the current literature is relatively limited compared with CTA; its higher cost and lower availability continue to restrict widespread clinical adoption.

Handheld Doppler is the most accessible option but is less sensitive than CTA (69% vs. 94%) and identifies fewer perforators per patient (10.4 vs. 14.8)^[26]. It is best suited for intraoperative and postoperative use rather than detailed preoperative planning.

DIRT is an emerging modality that detects cutaneous hot spots corresponding to perforators while providing angiosome-level perfusion information. Sensitivity (92%) is comparable to CTA (94%) and higher than handheld Doppler (69%), though it cannot assess intramuscular course and remains limited by a smaller evidence base^[2,26].

XR technologies - including virtual reality (VR), augmented reality (AR), and mixed reality (MR) - enhance visualization of existing imaging rather than generating new data. VR enables immersive 3D exploration with a strong correlation to CTA (R = 0.96)^[27]. AR allows intraoperative projection of perforator anatomy, reducing flap harvest time^[28]. MR platforms, such as Microsoft HoloLens, enable interactive evaluation of intramuscular anatomy^[29]. Its limitations include spatial registration error due to soft tissue deformation, with displacement greater than 3 mm in 30% of cases^[30]. Most studies evaluating XR technologies are early feasibility investigations with small sample sizes and limited clinical validation. At present, these technologies should be considered exploratory adjuncts rather than established components of routine DIEP planning. Although several AR, VR, and MR studies have been performed specifically in DIEP reconstruction, much of the broader literature remains limited to technical feasibility studies and other microsurgical applications.

No single modality fulfills all requirements for DIEP planning. CTA remains the primary modality with the strongest outcome data. MRA serves as an alternative when CTA is contraindicated. CDU provides real-time hemodynamic information and complements cross-sectional imaging. DIRT offers a radiation-free method for perforator localization, while XR technologies enhance spatial understanding and intraoperative navigation but depend on underlying imaging and technological quality. Optimal use is multimodal and tailored to institutional resources, expertise, and patient factors [Table 1]. Future comparative studies evaluating standardized imaging protocols and patient-centered outcomes will be important to better define the incremental benefit of combining multiple imaging modalities.

High-frequency linear-array probes (typically 5-15 MHz) are standard for perforator mapping^[1,31]. Higher frequencies (≥ 10 MHz) provide improved spatial resolution for superficial structures, enabling detection of vessels < 0.5 mm at the fascial level^[2]. However, increased frequency reduces penetration depth. In patients with thicker subcutaneous tissue, lower frequencies (5-8 MHz) may be required to visualize the intramuscular course of perforators within the rectus abdominis^[31]. Optimal settings for low-flow microvascular imaging include shallow depth (2-5 cm), low pulse repetition frequency (0.5-1.5 kHz or 3-10 cm/s), high color gain, and a low or disabled wall filter (~50 Hz)^[36]. These parameters improve the detection of small, slow-flow vessels while reducing motion artifact.

Multiple Doppler modes are used in perforator mapping. Color Doppler overlays directional flow information onto B-mode imaging, allowing real-time identification of perforator location and course^[2,3]. Power Doppler is more sensitive to low-velocity flow but does not provide directional information and may overestimate vessel diameter^[31]. In direct comparison, color-coded duplex sonography (CCDS) produced more accurate diameter measurements (mean 1.65 ± 0.45 mm) than power Doppler (1.31 ± 0.24 mm) when correlated with intraoperative findings^[37]. Spectral (pulsed-wave) Doppler enables quantitative waveform analysis, providing key hemodynamic data for perforator evaluation^[36,37]. Duplex ultrasound combines structural imaging with spectral Doppler, allowing simultaneous assessment of vessel anatomy and flow. Handheld continuous-wave Doppler detects flow but lacks anatomic localization and quantitative capability, with lower sensitivity (69%) compared to CDU^[26].

Spectral Doppler provides quantitative parameters relevant to perforator selection. Peak systolic velocity (PSV) reflects maximal flow and serves as an indicator of arterial inflow to the perforator. In perforator flap studies, mean PSV at the fascial level is approximately 17 cm/s, with a corresponding end-diastolic velocity (EDV) of approximately 5 cm/s and a resistive index (RI) of approximately 0.7, although reported values vary depending on flap type and vessel selection^[37]. A PSV threshold of ≥ 20 cm/s has been associated with successful perforator dissection in non-DIEP perforator flap studies, although DIEP-specific thresholds remain to be established^[38]. The RI, calculated as (PSV - EDV)/PSV, measures distal vascular impedance, with reported mean values around 0.7^[31]. The Pulsatility Index (PI) reflects resistance to flow and has demonstrated utility in postoperative monitoring, with elevated values associated with venous congestion and flap compromise^[39].

Microvascular flow imaging (MVFI) techniques improve the detection of slow-flow, small-caliber vessels beyond conventional Doppler^[40]. Technologies such as Superb Microvascular Imaging (SMI) and Micro Flow Imaging (MFI) use advanced clutter suppression to isolate low-velocity blood flow from tissue motion, enabling visualization of microvasculature without contrast^[40]. SMI has been used to assess vascularity in a small cohort of buried latissimus dorsi flaps after reconstruction^[40].

Contrast-enhanced ultrasound (CEUS) uses intravenously administered microbubbles that strongly reflect the sound waves to create bright, real-time images of blood vessels and perfusion. Together with 3D image rendering, it successfully identified all selected perforators in a retrospective series of 32 patients who underwent perforator flap reconstruction^[41]. However, most of these published studies evaluating MVFI have not been performed specifically in DIEP reconstruction, and their application to DIEP perforator mapping is currently inferred from broader microsurgical and vascular imaging experience. B-flow imaging has also demonstrated high sensitivity for perforator detection, although DIEP-specific validation remains limited^[42].

Ultrasound is applicable throughout all phases of DIEP flap reconstruction. Preoperatively, CDU enables identification, localization, caliber measurement, and hemodynamic assessment of perforators^[2,43]. Intraoperatively, it can confirm perforator patency and guide dissection when findings differ from preoperative imaging^[45]. Postoperatively, portable CDU allows serial monitoring of pedicle flow using the PI, which typically decreases over the first six postoperative days and provides an objective measure of perfusion, with elevated values associated with venous insufficiency (P = 0.03)^[39].

A tiered, patient-specific imaging approach leverages the strengths of each modality. In straightforward primary reconstructions, surgeon-performed CDU can be used as the initial mapping tool alongside established CTA or contrast-enhanced MRA modalities. In more complex cases - such as prior abdominal surgery, scarred tissue, or planned bipedicled flaps - CTA remains the preferred first-line modality due to its ability to define DIEA branching patterns, intramuscular course, and superficial venous anatomy with higher accuracy (95%-97%) than CDU (80%-84%)^[7]. In this framework, CDU serves as a complementary tool, providing real-time hemodynamic assessment and more accurate caliber measurement, while CTA supplies the broader anatomic overview^[6,7]. A practical summary of proposed imaging selection is provided in Table 2. Please note that this is not a validated algorithm or guideline.

CDU most directly impacts surgical decision-making in several scenarios. When CTA identifies multiple perforators of similar size, CDU-derived hemodynamic parameters, particularly PSV, can help identify the dominant vessel^[36]. CDU also provides more accurate caliber measurements than CTA, which tends to overestimate vessel diameter, thereby helping to avoid the selection of suboptimal perforators^[7].

In patients for whom CTA is contraindicated - such as those with contrast allergy, renal impairment, or concerns about radiation exposure - CDU may be a safe alternative with expertise^[7]. Additionally, when discrepancies arise between preoperative imaging and intraoperative findings - situations associated with increased flap loss - CDU enables immediate reassessment to guide perforator selection^[6].

Adoption of CDU is limited by operator dependence, a learning curve of approximately seven patients to achieve CTA-equivalent accuracy, lack of standardized training pathways, and limited large-scale randomized data^[1,4]. Although meta-analyses favor CTA in terms of flap loss and operative time, these findings are largely based on observational studies, while randomized data show no significant differences, suggesting that differences in operator experience and institutional expertise may contribute to the observed differences in outcomes^[1]. Beyond the learning curve itself, inter-operator variability remains an important limitation. Image acquisition, probe positioning, machine settings, and interpretation may differ substantially among operators and across institutions, depending on examiner experience, equipment, and scanning technique. Standardized scanning protocols and formal training pathways may improve reproducibility across institutions.

Artificial intelligence (AI)-assisted ultrasound analysis is an evolving area. Although early studies demonstrate encouraging proof-of-concept results, clinical validation in DIEP reconstruction remains limited, and prospective studies are needed before AI-assisted ultrasound can be routinely incorporated into DIEP reconstruction^[46]. Integration of ultrasound with AR platforms may eventually allow real-time projection of perforator anatomy onto the surgical field^[45]. However, most available studies remain technical feasibility reports rather than clinical outcome studies, and further prospective validation is required before widespread clinical adoption. Future research should prioritize prospective multicenter studies evaluating clinically meaningful outcomes rather than imaging accuracy alone, as improvements in perforator visualization do not necessarily translate into improved reconstructive outcomes.