DECT’s ability to define and identify diverse body contents based on material degradation has recently increased due to technological advancements. DECT can distinguish different materials because high-density substances, such as iron, calcium, or iodine, exhibit variable attenuation at different X-ray spectrum energy levels at two separate energy levels, while conventional CT scan use a single polychromatic x-ray energy to acquire images with a single peak energy of around 100–120 kV [6, 37]. VNCa-image reconstruction can remove calcium from cancellous bone using a three-material decomposition model [4, 38]. The VNCa image, produced from the data of bone mineral composition removal (comprising calcium), can display the underlying bone marrow. VNCa-image colour coding allows qualitative assessment, while region-of-interest measurements provide quantitative assessment [39]. Consequently, DECT permits the characterisation and distinguishing of various body structures using varied attenuation at different X-ray spectra and allows increased visualisation of bone-marrow-oedema distribution for identifying BMLs. The specificity of DECT was highly dependent on reader experience and the acquisition technique. For reader experience, the specificity of DECT in the diagnosis of TBML was dependent on the experience of readers. Although inter-reader agreement was excellent for DECT-VNCa images, board-certified experienced readers had a higher specificity of 96% compared with trainees (residents or research students) [3]. A larger voltage difference (at least 60) can result in higher sensitivity between low and high kilovoltage for spectral separation [3]. In addition, single-source consecutive DECT showed lower specificity compared with the dual source CT; moreover, fast kilovoltage switching techniques showed similar estimates of diagnostic accuracy as a dual source technique compared with dual-layer DECT [3]. Different studies used different Rel.CM values for diagnosing TBMLs [5, 6, 8, 39, 40]; their results showed that the VNCa image with a Rel.CM value of 1.45 could be an optimal parameter. Therefore, for the evaluation of TBML on DECT in the clinical, we should try to read the images by certified experienced readers instead of trainees. For the sensitivity and specificity, we slightly recommend to use dual source and dual-layer DECT with a larger voltage difference (at least 60) and Rel.CM value of 1.45 can be an appropriate parameter for VNCa post-processing procedure.

DECT successfully detects TBMLs after acute trauma of the spine, knee, wrist, hip, and ankle in patients with trauma events using the VNCa technique, according to numerous previously published studies [4, 6, 8, 41,42,43,44]. Li’s meta-analysis evaluated the accuracy of DECT-VNCa imaging in detecting TBMLs and revealed DECT’s high sensitivity and specificity for detecting TBMLs; DECT-VNCa imaging performs considerably well in diagnosing TBMLs [45]. DECT is a potentially important future diagnostic modality for evaluating TBMLs.

DECT’s advancement has allowed it to directly assess BMLs present in acute fractures through VNCa-technique development, which is comparable to that of MRI [5, 38, 41, 43,44,45,46,47,48,49] (Fig. 1). This potentially enhances the diagnosis of patients with traumatic and osteoporotic fractures.

Yang et al. conducted a meta-analysis of seven studies involving 510 vertebrae, evaluated DECT’s accuracy in detecting TBMLs in patients suffering from vertebral-compression fractures used MRI as the reference standard, and found that the sensitivity, specificity, and AUC value of DECT for detecting TBML were 0.82, 0.98, and 0.965, respectively [41]. Ghazi analysed 17 studies involving 2468 vertebrae in 742 patients with BML on MRI and found DECT to have a high sensitivity, specificity, and AUC, pooled the sensitivity, specificity, and AUC of DECT for vertebral body TBML were 89%, 96%, and 96%, respectively [3]. These findings suggest that DECT offers impressive diagnostic accuracy for TBMLs in vertebrae, resulted a moderate sensitivity and a high specificity for TBML identification, and suggested that positive predictive value is higher than negative predictive value. In the diagnostic performance of quantitative assessment, some studies have some contradictions [38, 43, 44, 48]. Comparing with the qualitative evaluation, the quantitative evaluation has a lower specificity in DECT for TBMLs. Some studies supported that the quantitative evaluation is more dependent on the experience of the readers, complex post-processing technology, and individual patient differences [44, 50]. However, the high diagnostic performance of qualitative assessment is more important because it is more feasible to perform in in trauma settings, instead of quantitative measurement that requires additional time and effort. Since DECT’s high specificity and positive predictive value, the detection of TBML in DECT can help to increase radiologists’ confidence to diagnosis. It allows the detection of BMLs associated with vertebral fractures and may obviate emergency confirmatory MRI. Furthermore, DECT allows TBML detection, even in patients without evident fractures. Therefore, DECT is a promising diagnostic tool for detecting vertebral TBMLs.

DECT exhibits similar diagnostic accuracy in depicting traumatic knee BMLs compared with vertebral TBMLs [8, 9, 39, 51,52,53,54,55] (Fig. 2). Previous TBML studies have concluded that DECT has high diagnostic accuracy, which is comparable to that of MRI.

Knee TBMLs. NM Not mentioned

According to a meta-analysis that used MRI as the reference standard, DECT demonstrates excellent specificity in evaluating TBMLs in patients with acute knee injuries [54]. However, DECT exhibits slightly lower sensitivity in detecting TBMLs than MRI, indicating that when DECT yields negative results in patients presenting symptoms, MRI may be required to further detect occult BMLs [54]. Moreover, several studies that have evaluated patients with acute knee injuries using DECT and MRI suggested that DECT is specific and accurate in detecting TBMLs in adult patients with acute knee injuries. DECT-VNCa reconstructions exhibit superb diagnostic performance in terms of sensitivity, specificity, and accuracy in depicting TBMLs compared with MRI [8, 39, 52, 53]. Furthermore, Booz’s study found DECT colour-coded map visualisation of TBMLs to have great diagnostic accuracy for TBML characterisation owing to its high-resolution and bone window compared with conventional CT images [39]. Moreover, in Foti’s study, DECT images improved the overall accuracy of TBML visualisation in chronic knee injuries [9]. These results confirmed DECT’s favourable diagnostic accuracy in depicting TBMLs around the knee joint both in acute and chronic traumatic situations compared with that of MRI. Ultimately, these results corroborate DECT’s excellent specificity and accuracy in detecting TBMLs in adult patients with suspected bone fractures or symptoms after acute knee injury, especially when MRI is contraindicated or unavailable.

Recent studies have primarily investigated the DECT-VNCa technique’s diagnostic ability in detecting TBMLs in the appendicular skeleton, such as in the ankles, wrists, calcaneus, and lower limbs, compared with that of MRI [6, 42, 46, 56,57,58] (Fig. 3). They evaluated DECT’s diagnostic accuracy in detecting TBMLs throughout the appendicular skeleton. In a meta-analysis by Wilson, TBML detection in the appendicular skeleton by DECT was found to have excellent sensitivity, specificity, and AUC values [59]. According to Guggenberger’s study, DECT-reconstructed non-calcium images can help diagnose separate ankle-joint TBMLs with great sensitivity and negative predictive value and intermediate specificity and low positive predictive value [6]. Two studies demonstrated DECT’s utility in reliably identifying BMLs in the trauma wrist [42, 57]. DECT coupled with the VNCa algorithm was used to detect BMLs in wrist fractures and demonstrated reliability and user friendliness, rendering the DECT-VNCa algorithm a promising BML-detection technique in acute carpal trauma, with potential to improve CT’s diagnostic value in this injury type [42, 57]. Another study demonstrated the clinical applicability and high diagnostic accuracy of colour-coded DECT-VNCa reconstructions in the visualisation of calcaneus TBMLs compared with that of MRI [46]. This study demonstrated colour-coded-VNCa reconstruction’s comparability with MRI regarding diagnostic confidence and image noise, revealing no significant difference [46]. Accordingly, DECT exhibited high diagnostic accuracy and confidence in detecting TBMLs in patients with acute trauma injury to the appendicular system. Where MRI is unavailable or contraindicated, DECT can be performed as a feasible alternative for patients with acute trauma and probable BMLs, potentially leading to more efficient patient care after appendicular-skeleton trauma.

Appendicular-skeleton (excluding knees) TBMLs. NM Not mentioned