Fig. 1.6  ADC maps (units 10−6 mm2/s) in a patient with rectal tumour obtained by signal fitting using b = 200, 500 and 1000 s/mm2 (upper image) and by signal fitting using b = 0, 200, 500 and 1000 s/mm2 (lower image). The tumoural ADC value measured on the lower image comparing to the higher image because it is influenced by the large micro-perfusion contamination that is evident when using low b value data in our mono-exponential fitting algorithms

1.3\ Clinical Applications

According to ESGAR recommendations [22], use of diffusion-weighted imaging (DWI) in rectal cancer patients is not obligatory for primary baseline staging. DWI sequences, however, are implemented in routine rectal cancer MRI protocols nowadays. In tissues presenting with high cellular content and intact cell membranes, such as rectal tumour tissue, water motion is restricted. This does not apply to the tissue from which tumour originates, in our case rectal wall, and thus DWI is very

Fig. 1.7  Patient with rectal cancer, axial T2-weighted TSE (upper left), calculated b 1500 (upper right), ADC (lower left) and originally acquired b 1500 (lower right). Calculated b value image successfully demonstrates the high signal intensity of the tumour with better SNR as compared to originally acquired b 1500 image

Fig. 1.8  Whole tumour histogram analysis in a patient with rectal cancer. Histogram metrics are shown together with the tumour histogram and can be used to assess treatment-related changes (i.e. chemoradiation effects)

helpful in tumour detection. DWI sequences for clinical purposes are relatively quick to perform, do not require the administration of contrast medium, and can be integrated to the existing rectal tumour imaging protocol without significant delay in the overall examination time. Optimally, DWI images have to be obtained with the same slice thickness and orientation as the high-resolution axial oblique T2-weighted images. For quantitative analysis or measurements for assessment of therapy response, at least three b values should be obtained [23]. It is advised as there is no standardization regarding the use of DWI in the abdomen, and more specifically for rectal cancer imaging, each centre should test the apparent diffusion coefficient (ADC) values produced by their sequence or scanner and have own references.

Lack of DWI protocol standardization for post-CRT MRI leads to variability in ADC values measured, with no universal cut-off value for differentiating between viable tumour and complete response (CR). However, post-neoadjuvant treatment means ADC values of responding tumours have a tendency to increase. Using an axial-orientated diffusion-weighted sequence with background body signal suppression and b values, 0, 500, and 1000 s/mm2, Curvo Semedo et al. [24] demonstrated that lower ADC values were associated with a more aggressive tumour profile. ADC has the potential to become an imaging biomarker of tumour aggressiveness profile. Kim et al. [25] reported that the diagnostic accuracy in evaluating CR significantly increases when DWI is combined with conventional MRI. However, restricted diffusion in the corresponding tumours has been reported in 42% of the patients who achieve pathological complete response (pCR) after neoadjuvant CRT, reasons for this according to surgical specimen pathology review being the presence of intramural mucin, degree of proctitis and mural fibrosis [26].

1.3.1\ Whole-Body Diffusion

Whole-body MRI is already incorporated in the clinical routine due to recent technological advances in the field of RF technology including integrated high-density phased array coils that can cover large anatomical areas, providing an adequate signal- to-noise ratio. Integrated parallel imaging techniques accelerated image acquisition and resulted in significant reductions in examination times making whole-body MRI clinically feasible. The main area of clinical applications of whole-body MRI is screening for metastatic disease [27]. The workhorse sequence to detect metastatic disease is diffusion-weighted imaging due to the very high conspicuity that offers to detect hypercellular lesions. A major challenge in terms of image quality on diffusionweighted images is efficient fat saturation that can be very difficult to achieve in large anatomic areas. The latter can be addressed to the presence of air in the gastrointestinal tract and the variable amount and distribution of fat due to different body habitus. Therefore, the mainstream pulse sequence that should be recruited for whole-body diffusion applications should provide robust fat suppression. Such a method should make use of inversion pulses to achieve signal nulling of the fat, the so-called STIR technique (short tau inversion recovery), which is well known to be less sensitive to Bo inhomogeneity. The diffusion sequence making use of inversion pulses to achieve fat saturation is called DWIBS (diffusion-­weighted imaging with body suppression) and is usually acquired in axial plane with a single b value ranging between 600 and 1000 depending on the gradient performance of the MRI scanner.

Conclusions

Diffusion-weighted imaging of the gastrointestinal tract is technically challenging due to the presence of multiple physiologic motions (respiration, peristaltic motion), as well as the presence of air. With careful optimization of pulse sequence parameters, it has the potential to aid radiologists not only on differential diagnosis but also on assessing and predicting treatment response.