Image quality in medical imaging is influenced by many factors, with one crucial factor being the strength of the magnetic field of the scanner. The 3T scanners offer a fundamental advantage over 1.5T scanners due to their higher signal-to-noise ratio (SNR). The SNR is a parameter that theoretically increases linearly with the static magnetic field [19]. Consequently, 3T scanners can potentially provide enhanced image quality compared to their 1.5T counterparts.

It is worth noting that both 1.5T and 3T scanners can yield satisfactory and reliable diagnostic results when the acquisition parameters are optimized and state-of-the-art technology is employed [7]. According to the PI-RADS v2.1 guideline, both 1.5T and 3T scanners are considered suitable for performing prostate MRI; however, most experts prefer and recommend 3T [20]. In a study conducted by Ullrich et al. [21], a comparison was made between the 1.5T and 3T MRI systems, revealing that the SNR and the “contrast-to-noise ratio” (CNR) for T2-weighted images (T2WI) exhibit similar performance across both field strengths. However, when it comes to diffusion-weighted images (DWI), the SNR and CNR notably diminish at 1.5T. Given the particular significance of DWI in identifying csPC within the peripheral zone (PZ), the preference lies with the utilization of 3T scanning. Exceptions exist, and there are some situations in which it is preferable to perform the examination on a 1.5T scanner, namely:

- When patients have implanted devices, it may not be safe to perform the exam in 3T.

- While patients can undergo MRI safely using 3T scanners, it is important to consider the presence of implanted devices, such as bilateral metallic hip prostheses. In some cases, the proximity of these devices to the area being imaged can lead to artifacts that negatively affect image quality. Therefore, caution should be exercised when imaging patients with such implants, to ensure optimal image interpretation.

It is essential to emphasize that conducting prostate MRI at lower magnetic field intensities, particularly, those that are below 1.5T, is not advisable and is currently not being practiced [20].

The limitations inherent to mpMRI can be classified into two categories: factors affecting image acquisition quality and those influencing accurate interpretation [22].

To align with PI-RADS 2.1 recommendations, mpMRI testing requires a minimum system field strength of 1.5T [20].

Crucial sequences within mpMRI, such as DWI and dynamic sequences acquired with intravenous administration of paramagnetic contrast medium (DCE) imaging, extend beyond morphological assessment. These sequences are notably sensitive to motion artifacts induced by prostate spasms, muscle movements, and even intestinal motility. To counter this, implementing strategies to limit intestinal motion is suggested (check section 2.5).

Inherent patient attributes can impose additional constraints on prostate MRI [22]. The strength of organ signals, a crucial determinant of image quality, depends on their proximity to the MRI receiver coil. In cases of severe obesity, where increased adipose tissue thickness creates a greater distance between the coil and the prostate, image quality may suffer a significant decline. Furthermore, the presence of metallic foreign objects, especially hip endo-prostheses, which tend to become more common with age, can introduce substantial limitations. Metal endo-prostheses can induce field distortions, potentially compromising the reliability of mpMRI assessments [22].

To achieve the best outcomes, it is imperative that mpMRI studies be reviewed by experienced radiologists [23]. These studies are most effectively conducted within specialized reference centers, thereby augmenting quality through the input of multiple experienced professionals. The proficient interpretation of morphological data demands a thorough patient profile, ideally provided in a comprehensive manner by urologists. This collaborative effort between radiologists and urologists cultivates mutual expertise, ensuring that radiologists furnish essential information for urologists, who, in turn, adeptly decipher radiological descriptions [23].

Prior to undergoing an MRI, thorough contraindication screening is imperative to ensure patient safety. Essential checks encompass implants, metallic foreign bodies, contrast allergies, and renal function. In cases of potential contraindications, consultation with a radiologist is prudent. Guidelines for contrast agent contraindications and renal function are outlined in the European Society of Urogenital Radiology (ESUR) contrast media guidelines [24].

Post-biopsy hemorrhage, contrary to earlier assumptions, does not hinder mpMRI. A study by Rosenkrantz et al. [25] demonstrated that substantial hemorrhage and a brief post-biopsy interval do not compromise the accuracy of csPC detection via mpMRI. Despite this finding, the latest European Association of Urology (EAU) guidelines recommend prioritizing MRI before biopsy [26].

In patients with contraindications to mpMRI, alternatives for local staging encompass transrectal ultrasound and abdominopelvic CT [26]. However, both methods have limited accuracy for local staging [27]. In cases of locally advanced cancers, abdominopelvic ultrasound or CT might reveal rectal or bladder invasion, along with upper collecting system dilatation [27].

The utilization of mpMRI has revolutionized the diagnostic approach to PC. It serves not only as an imaging technique for identifying suspicious lesions but also as an intraprocedural guide for targeted biopsies (MRI-TBx), thereby effectively closing the diagnostic gap [28].

Multiple randomized trials conducted at both single and multicenter levels, including studies, such as PROMIS [29], MRI-first [30], PRECISION [31], and the investigation by van der Leest et al. [32], have consistently demonstrated a superiority by the “MRI pathway”. This approach involves employing mpMRI and MRI-TBx on suspicious lesions, outperforming the conventional “standard pathway” of the standard biopsy for all patients with elevated PSA levels and/or positive digital rectal exploration results. MRI-TBx offers the advantage of reducing the overall number of biopsies performed while also diminishing the diagnosis of clinically insignificant diseases. Simultaneously, it either maintains or enhances the diagnosis of csPC [28].

Recent studies have revealed that using an endorectal coil (ERC) in prostate imaging does not necessarily lead to increased tumor detection rates. However, employing an ERC can offer the advantage of obtaining higher-resolution images and, as a result, may improve local staging. However, although the ERC enhances image quality, it is also associated with several drawbacks. These include deformation of the gland contour, enlargement of the near-field coil, increased cost and examination time, and greater patient discomfort [33, 34]. Currently, with the continuous improvements in hardware and software, it is possible to obtain adequate and accurate images without it [19]. Consequently, the use of ERC is no longer recommended as routine [34].

The PI-RADS v2.1 guideline does not provide a definitive consensus on patient preparation for prostate MRI. In cases where an ERC is not utilized, the presence of air and/or stool in the rectum can cause artifactual distortion, potentially compromising the quality of the diffusion-weighted imaging (DWI). To mitigate this issue, a minimal preparation enema, administered by the patient prior to the examination, may prove beneficial, especially when an ERC is not used. However, it is important to note that an enema can stimulate bowel peristalsis, leading to increased motion-related artifacts in certain instances [20].

In some patients, the use of an antispasmodic agent may help reduce these motion artifacts from the bowel peristalsis. However, it is not always necessary, and factors such as potential adverse drug reactions and the additional cost should be carefully considered before administering [20].

A point of general agreement is that patients should, whenever feasible, empty their rectum just before undergoing an MRI examination. This is because rectal distension caused by a stool or gas has been proven to markedly amplify the extent of geometric distortion within the prostate gland [20].

A comprehensive prostate mpMRI protocol should incorporate a combination of morphological sequences, including T1-weighted imaging (T1WI) and T2-weighted imaging (T2WI) with two functional sequences: DWI and DCE [35]. Each sequence serves a specific purpose in evaluating prostate pathology.

T2-weighted sequences play a pivotal role in prostate MRI, serving as the cornerstone of the imaging protocol. It is crucial to acquire T2WI in the axial plane, and in at least one orthogonal plane, such as sagittal or coronal, to facilitate the morphologic assessment of lesions [20]. These images provide optimal visualization of the zonal anatomy of the prostate, enable the assessment of abnormalities within the gland, and aid in determining the presence of ECE, SVI, and LN spread [19].

Conversely, T1-weighted images are primarily utilized to identify the presence of a hemorrhage within the prostate and seminal vesicles, to delineate the gland’s contour, and as a potentially valuable tool in detecting skeletal and nodal metastasis, particularly following the intravenous administration of paramagnetic contrast medium [20].

An essential component of mpMRI is DWI, a sequence that reflects the random Brownian motion of water molecules within tissues. This functional sequence should include both high b-value images and the apparent diffusion coefficient (ADC) map [20].

The degree of cellularity in PC correlates with GS, which reflects the aggressiveness of the cancer. As the GS increases, there is a greater reduction in the water diffusion capacity. In simpler terms, PC demonstrates restricted diffusion of water molecules, with higher GS presenting more significant restrictions [36, 37].

Dynamic contrast enhancement involves acquiring T1WI before and after the intravenous administration of a gadolinium-based contrast agent (GBCA). It is recommended to apply fat suppression and/or subtraction techniques to improve image evaluation [20].

Similar to other types of tumors, PC typically exhibits early enhancement in the dynamic study compared to normal tissue. A positive finding on DCE is characterized by focal enhancement that occurs earlier or simultaneously, in relation to the enhancement of adjacent normal prostatic tissues. This enhancement pattern typically correlates with suspicious findings by T2WI and/or DWI [20].

During the dynamic study of mpMRI, it is crucial to conduct a thorough evaluation, specifically focusing on the identification of early focal enhancement. If such enhancement is detected, it is essential to carefully examine the corresponding T2WI and DWI to assess the associated abnormality [20].

Currently, the additional value of DCE in PC is not definitively established. Most published data indicate that the incremental benefit of DCE, when added to the combination of T2WI and DWI, is modest. Therefore, while DCE remains an integral component of mpMRI, its role in determining the PI-RADS v2.1 assessment category is considered secondary to T2WI and DWI [20].

Technical specifications in the prostate mpMRI protocol, according to the PI-RADS v2.1 guideline, and some considerations regarding the protocol for the use of a 3T scanner at the authors’ institution are detailed in Table 1.

According to EAU guidelines, local staging is recommended specifically for intermediate and high-risk groups of patients (PSA $\ge$10 ng/mL or GS $\ge$7 or cT2b or higher clinical stages). The primary goal of local staging, in these cases, is to differentiate between diseases confined to the prostate and locally advanced diseases [2].

T3a is characterized not only by the presence of ECE but also by the neurovascular bundle involvement, as well as microscopic invasion of the bladder, neck, and internal bladder sphincter [2]. Histopathologically, ECE is further subclassified into focal and established subtypes [43]. Focal ECE is identified when there are only some foci outside the prostate, and this subtype generally carries a better prognosis compared to established ECE. However, the detection of focal ECE using mpMRI has certain limitations. When there are multiple foci present, it is considered an established ECE, which is associated with a worse prognosis, particularly if there is a positive surgical margin. In such cases, the accuracy of mpMRI is higher [44].

Expert readers are required to evaluate the possibility of ECE [45]. The European Society of Urogenital Radiology (ESUR) has suggested a scoring system to predict ECE, known as the ESUR score, which provides qualitative descriptors designed to assess ECE (Table 3) [46].

The probability of an extracapsular extension (ECE) is evaluated using a five-point scoring system that assigns an ordinal risk score to signify the likelihood of an ECE. It is crucial to highlight that this scale does not incorporate a score of two.

In a study conducted by Boesen et al. [47], in 2015, the authors scrutinized this scoring model and identified that a cutoff threshold of $\ge$4 yielded the most favorable equilibrium between sensitivity (0.81) and specificity (0.78).

Furthermore, the length of the tumor contact with the capsule, as observed by MRI, has been shown to be a strong predictor of ECE, by demonstrating good to excellent agreement among different readers [48, 49, 50]. Nonetheless, it is important to acknowledge the considerable diversity in the threshold values documented in various studies, spanning from 6 to 20 mm. In response to this incongruity, the PI-RADS v2.1 guidelines have incorporated an arbitrary reference point, setting the cutoff at a length of 10 mm [49].

A recent development in this field is the introduction of a novel grading system by Mehralivand et al. [51], known as the extraprostatic extension (EPE) grade, which utilizes MRI features to predict the likelihood of an ECE. The grading system defines the following categories:

- Grade 0: No suspicion for EPE.

- Grade 1: Either tumor contact length $\ge$15 mm or there is capsular bulge/irregularity.

- Grade 2: Both features defined for grade 1 are present.

- Grade 3: Visible ECE is detected by MRI [51].

In 2022, Caglic et al. [52] evaluated the capsular enhancement signal (CES) role in DCE imaging to detect ECE. Their findings demonstrated that CES is a highly predictive indicator of ECE presence and is closely associated with increased tumor aggressiveness [52]. Moreover, they conclude that the presence of CES may offer significant clinical utility in predicting ECE, particularly in the small but significant percentage of patients where this signal is found. This observation can be visually appreciated in Fig. 10, which illustrates the distinctive CES characteristic on MRI scans (Fig. 10).

To evaluate the seminal vesicles effectively, a combination of T2WI and functional imaging is recommended. Recent studies have proven that incorporating functional sequences, particularly DWI, enhances the accuracy of seminal vesicle assessment more significantly than DCE [55, 56]. In addition, acquisition/reconstructions in the sagittal and coronal planes play a vital role in characterizing the pattern of tumor dissemination, whether it is through the ejaculatory ducts (type I) or by ECE (type II) [7].

Hypointensity on T2WI, DWI restriction, or DCE enhancement in the seminal vesicle region indicates SVI (Figs. 11,12). Therefore, it is essential to always evaluate the three orthogonal planes and look for asymmetries.

Finally, T4 PC involves the invasion of adjacent structures beyond the seminal vesicles, including the bladder (Fig. 13), external urethral sphincter, levator ani muscles, pelvic wall, and/or rectum (Fig. 14).

Given its poor prognosis and elevated risk of biochemical recurrence, most guidelines recommend a combination of androgen deprivation therapy and radiation to manage T4 PCs [58]. Consequently, the accurate detection of a T4 disease before surgery is of paramount importance to ensure appropriate treatment planning.

Nodal spread is seen in 5–10% of PCs at the time of radical prostatectomy [59]. Prostate cancer primarily metastasizes to four specific nodal stations in the pelvis, freferred to as regional nodes: the obturator chain, internal and external iliac chains, and presacral chain [60]. The involvement of any regional node is classified as stage N1. Conversely, the involvement of non-regional stations (e.g., paraaortic, paracaval, deep inguinal, femoral, and common iliac lymph nodes) represents the M1a disease (Fig. 15) [61, 62].

A retrospective study conducted by Asfuroğlu et al. [53], in 2022, sought to determine the most effective MRI-based method for predicting EPE. They compared the performance of the ESUR score, Likert scale (a score based on a subjective overall assessment of a combination of mpMRI criteria by an experienced radiologist), tumor contact length, and EPE grade. The study concluded that all four methods exhibited good diagnostic performance in predicting EPE. However, the study specifically highlighted the EPE grade as a promising approach due to its utilization of both qualitative and quantitative parameters. Moreover, the EPE grade was found to be less dependent on the reader’s experience, further enhancing its reliability as a predictive tool for EPE [53].

Considering the lack of a definitive consensus on ECE-related signs for PC, further studies are needed to establish the most effective methods for predicting ECE when utilizing mpMRI.

In addition to detecting ECE, pinpointing the specific location of T3a is vital since clear surgical margins become harder to achieve in the apex. Conversely, tumors situated away from the neurovascular bundle, e.g., anteriorly, allow for nerve-sparing surgery, thereby reducing postoperative complications, such as incontinence and erectile dysfunction [54].

T3b disease involves neoplastic growth in one or both seminal vesicles. This occurs through PC cell penetration via ejaculatory ducts or direct contact with the prostatic capsule. SVI prevalence ranges from 4% to 23% [55, 56].

Ejaculation abstinence aids SV distension; however, the PI-RADS v2.1 benefits are uncertain [20].

Additionally, T3b patients face heightened risks of LN involvement, local recurrence, and distant metastasis. Identifying SVI preoperatively becomes crucial for prognosis and treatment planning [57].

Differentiating between T3a and T3b stages is of utmost importance when determining the appropriate therapeutic approach. Thus, patients with T3b tumors are more likely to benefit from external beam radiotherapy or androgen deprivation therapy instead of radical prostatectomy or brachytherapy [7].

Studies mapping the distribution of nodal metastases demonstrated that around 75% of pelvic nodal metastases are located within the obturator chain and the internal and external iliac chains, while the remaining 25% are found in the aortic bifurcation, common iliac chain, and presacral chain [60].

The PI-RADS v2.1 guideline, in alignment with the EAU recommendations, emphasizes the importance of performing N staging using mpMRI in all patients, irrespective of their risk classification, and even prior to the biopsy. This approach entails not only the employment of standardized sequences, to distinguish confined neoplasia from locally advanced neoplasia (T staging), but also the utilization of an appropriate FOV that extends up to the aortic bifurcation. This expanded FOV allows for the evaluation of common iliac lymph nodes and bifurcations (M1a), to facilitate the partial staging of the bony pelvis (M1b) [35].

Currently, the identification of abnormal LN(s) by MRI, primarily, relies on assessing their size, morphology, shape, and enhancement pattern [20]. Some of the suggested evaluation criteria include a size exceeding 8 mm, a round shape, the loss of the fatty hilum, a low T2WI signal relative to the primary tumor, and an irregular border (Fig. 16) [7].

Size alone is a useless criterion for evaluating a positive LN disease. Not only is there significant dimensional overlap between positive and negative LN(s) but certain reactive inflammatory changes can also produce false positive results. In addition, smaller LN(s) may harbor micrometastases that do not affect the LN size, thus, leading to incorrect staging [63].

Recent studies have revealed that a significant proportion of positive lymph nodes dissected during ePLND, after MRI, have a short axis smaller than 5 mm, below the conventional cutoff point of 8 mm [64].

Despite advancements in MRI, the current diagnostic performance for nodal staging remains suboptimal. Consequently, ePLND remains the gold standard procedure for the accurate detection of LN metastases in PC [26].

More work and development are needed to identify an imaging technique that allows for a more efficient and less invasive assessment of patients with metastatic LN(s). Indeed, initial clinical trials of a prostate-specific membrane antigen ${}^{68}$Ga-Prostate-Specific Membrane Antigen (PSMA) Positron Emission Tomography (PET)/CT and PET/MRI have shown promising results in the detection of LN metastases, leading to treatment plan adjustments (including systemic treatment) [65, 66].

According to the EUA guidelines [26], patients categorized as intermediate- and high-risk should undergo CT imaging of the abdomen and pelvis to evaluate non-regional LN (M1a disease) and assess visceral metastasis (M1c disease). Additionally, bone scintigraphy is recommended for evaluating bone metastasis (M1b disease).

In recent years, whole-body MRI (WB-MRI) has gained attention due to its ability to detect bone marrow infiltration by malignant cells before visible bone remodeling occurs (which can be detected by bone scintigraphy) [67]. For this reason, WB-MRI is particularly useful in M staging, especially in M1b cancers.

The METastasis Reporting and Data System for Prostate Cancer (MET-RADS-P) provides practical guidance for the acquisition, interpretation, and reporting of WB-MRI in advanced PC [68]. The PI-RADS guidelines for the evaluation of metastasis in WB-MRI recommend a protocol consisting of a combination of anatomical (T1WI, short tau inversion recovery [STIR] or T2WI with fat suppression) and functional (DWI) sequences [68].

Despite the diagnostic potential of WB-MRI, several limitations hinder its widespread use for PC staging. The main challenges include high costs and long scanning times [7].