Abstract

Introduction

Since its introduction into medicine over 50 years ago, the technology used in CT scanners has undergone many dramatic changes, enabling single breath-hold scanning of the body, stop-motion imaging of the heart, and quantitative imaging using dual-energy techniques. Energy-resolving, photon-counting detector (PCD) CT is the latest advance in CT technology. It differs from all earlier CT systems, which used energy integrating detectors (EIDs), in the way that X-rays are converted to electronic signals. Among the advantages of PCD-CT is increased dose efficiency, which permits a reduction in radiation dose relative to EID-CT while maintaining or improving image quality. ^1–5 Additional benefits include improvements in spatial resolution, increased iodine signal, and reduced electronic noise. ^6–16

The first PCD-CT system to become commercially available (NAEOTOM Alpha, Siemens Healthineers, Forchheim, Germany) utilizes a dual-source geometry. ⁷ Because children are generally more prone to movement during scan acquisition than are adult patients, pediatric CT protocols tend to prioritize rapid image acquisition to reduce motion artifacts, in addition to radiation dose reduction. ^17–21 Dual-source CT offers one method to optimize scan speed, where two X-ray tube/detector pairs scan overlapping body regions at high table speeds, achieving a temporal resolution that is a factor of two superior to single-source geometries. ²² These dual-source, high-pitch (up to 3.2) “Flash” scans have the potential for subsecond scans times (depending on body region) and per image temporal resolution as fast as 66ms. ²³ While these speeds match current dual-source EID-CT capabilities, the use of PCDs allows for simultaneous multienergy imaging, providing the ability to perform multienergy imaging with high-pitch, high-temporal resolution or ultra-high spatial resolution. ^7,24–27 Table 1 offers sample protocols for pediatric patients that demonstrate these advantages. This combination of capabilities optimizes CT image quality in a manner that is beneficial for pediatric patients by combining simultaneous rapid imaging acquisition, multienergy capabilities, and improved spatial resolution for imaging small anatomic parts. ^28–30

Table 1.

Sample protocols

	CARE Dose4D & CARE keV	CARE keV optimized for	CARE keV IQ level	Effective mAs	kV	Pitch	Rotation time (s)	Scan modea (short name)	Collimation (mm)	Kernel
Chest without contrast	Manual	Non-contrast	35–45	Varies	120	3.2	0.25	Flash+SR+ME or Flash+UHR	144 × 0.4 (SR) or 120 × 0.2 (UHR)	Br44, Qr56
Chest with contrast	Manual	Soft Tissue	35–45	Varies	120	3.2	0.25	Flash+SR+ME	144 × 0.4	Br44, Qr56
Chest without contrast (low dose)	Manual	Non-contrast	2–6	Varies	100Sn	3.2	0.25	Flash+UHR+Sn	120 × 0.2	Br44, Qr56
Abdomen Pelvis with contrast	Manual	Soft Tissue	120–140	Varies	120	3.2	0.25	Flash+SR+ME	144 × 0.4	Br44, Qr44
Abdomen Pelvis without contrast	Manual	Non-contrast	120–140	Varies	120	3.2	0.25	Flash+SR+ME	144 × 0.4	Br44
Extremity without contrast	Off	Off	Off	65–160	120	1	0.5–1	UHR+ME	120 × 0.2	Br76 to Br84
Extremity CTA	Manual	Vascular	120–140	Varies	120	0.5	0.25–1	UHR+ME	120 × 0.2	Bv38 to Bv60
Temporal Bone without contrast	Manual	Non-Contrast	220	Varies	120	0.5	1	UHR+ME	120 × 0.2	Hr44, Hr84
Abdominal CTA	Manual	Vascular	120–180	Varies	120	0.5	0.25	UHR+ME	120 × 0.2	Bv38 to Bv60

Open in a separate window

^aSR, standard resolution; ME, multienergy-energy; UHR, ultra-high resolution

This article will describe PCD-CT, detail the differences between PCD-CT and EID-CT, and illustrate current and possible future clinical applications in children.

Unique features and benefits of photon-counting CT detectors

EID-CT uses solid-state scintillators made of materials such as gadolinium-oxysulfide, which convert X-ray photons to visible light, which is subsequently converted to electrical signal by a photodiode (Figure 1a). ^{7,9–11,25,31,32} The detector elements (detector pixels) are separated by reflective septa to reduce the spread of visible light to adjacent detector elements. The output signal of each detector element is the summation of the energies of all detected photons, where higher-energy photons are weighted more heavily in the total output signal. ^9,10 Lower energy X-ray photons contain more inherent contrast information because they interact more frequently by the photoelectric effect, which enhances contrast in biological tissues. Low-energy photons above 33 keV are also closer to the k-edge of iodine, a point at which iodine signal strongly increases. However, in EID-CT, lower-energy photons represent a lower percentage of the total output signal from the detector, lowering the contrast-to-noise ratio (CNR) of soft tissues, bone, and iodine. ^9,10

Open in a separate window

Figure 1.

(a) Schematic of an EID. EID-CT uses a scintillator to produce visible light when an X-ray photon strikes it; this light is then converted to electrical signal by a photodiode. Reflective septa are positioned between detector elements to minimize the spread of light to adjacent detector elements. (b) Schematic of a PCD. PCD-CT directly generates charge pairs (electrons and holes) using a semi-conductor, with the charges traveling across a potential difference to induce an electrical signal at the pixelated anodes. (c) A low-energy threshold (e.g. 20 keV) is utilized to eliminate electronic noise, which lies below that energy. A high-energy threshold (e.g. 65 keV) can also be set to divide the detected photons into low- and high energy data sets. EID, energy integrating detector; PCD, photon counting detector.

PCD-CT utilizes semi-conductors such as cadmium zinc telluride (CZT), cadmium telluride (CdTe), or silicon (Si), ³² which enable the direct conversion of X-rays to electrical signal, without the conversion to visible light. ^9–11 The theoretical feasibility of pediatric applications for photon-counting CT were first demonstrated using a Monte Carlo model with silicon detectors by Yveborg et al in 2009. ³³ The current commercially available scanner (NAEOTOM Alpha; Siemens Healthineers) utilizes CdTe. ^7,9,32 When an incident X-ray photon interacts with the semi-conductor material, it creates a cloud of electrons and holes, which then move across a high voltage potential to create 10–20 ns electrical pulses at the pixelated anodes (Figure 1b). ^9–11,34,35 These pixelated anodes form the individual detector elements. The anodes are bonded to an application-specific integrated circuit (ASIC) that provides parallel read-out channels on the back side of the detector. ⁹ At higher incident photon energies, more electrons and holes are created and the signal amplitude is increased. A pulse height analyzer subsequently compares the pulse height against preset thresholds to determine which counter should be incremented (Figure 1c). ¹⁰ In this manner, the energies of different photons are resolved, rather than summed.

System development

The development of PCD-CT has occurred over the time span of nearly 20 years and has included the creation of various small animal and human research systems. The primary technical challenge has been to fabricate in a robust and affordable manner detectors capable of operating at CT photon flux levels (up to 10⁹ photons/s/mm²). Investigational PCD-CT systems developed over the past decade include two whole-body research systems (SOMATOM Count and SOMATOM Count Plus, Siemens Healthineers), both using CdTe detectors. In two other research systems, a CZT-based detector is used (Philips Healthcare, Haifa, Israel) ^36–38 and a Si-based detector is used (GE Healthcare, Chicago, IL). ³⁹

In 2021, a clinical high-flux-capable PCD-CT system (NAETOM Alpha) became commercially available. ^7,25 It offers a range of tube potentials (70, 90, 120, 140 kV) and tube currents between 10 and 1300 mA. ²⁵ Tin filtration is available at 100 and 140 kV. Two detector collimations are available: the standard 140 × 0.4mm (5.76cm coverage per rotation) and the ultra-high resolution 120 × 0.2mm (2.4cm coverage). Several recent publications detail the specific scan protocols used in pediatric patients. ^4,28–30

Spatial resolution improvement

The design of PCDs is such that reflective septa are not required, as they are in EIDs. This improves the geometrical dose efficiency because septa can fill about 20–30% of an EID’s surface area, absorbing photons that have passed through the patient without contributing to image signal. ¹⁰

To improve the spatial resolution of a CT system, the detector elements must become smaller. In EID-CT, this would further increase the “dead space” occupied by septa as a percentage of the detector’s surface area. To achieve better resolution without taking this step, which would impose a dose penalty across all scan types, some scanners have the option of moving an ultra-high resolution attenuating grid or comb “filter” in front of the detector to decrease the effective size of the detector element by blocking incident X-rays. ⁴⁰ This imposes a radiation dose penalty of approximately a factor of 2, and hence is used in only limited applications, such as imaging of the inner ear or the extremities. Given the emphasis on radiation dose reduction in pediatric CT protocols, ^17,18,41 it is important to note that the detector design of PCD-CT eliminates the need for grid or comb filters. ^6,8,16

In the currently available PCD-CT, ultra-high spatial resolution can be achieved solely with the use of very small detector elements (0.151mm by 0.176mm), without the need for additional attenuating grids or combs, allowing creation of slice thicknesses as narrow as 0.2mm without incurring a dose penalty. ^7,9,11 Ultra-high spatial resolution imaging of the torso, such as for shoulders and pelvis imaging is therefore possible with this system. ^16,42

The spatial resolution of a CT system is measured by the modulation transfer function (MTF), which describes the ability of the system to accurately convey information as a function of spatial frequency (in line pair per cm). The current commercial PCD-CT has a 10% MTF value of 36.1 lp/cm, compared to 19.6 lp/cm on a high-resolution (0.25mm detector element) EID system. ^7,43 The limiting spatial resolution (0% value of the MTF) of the NAETOM Alpha is reported to be 40 lp/cm, which can resolve high contrast objects of approximately 0.125mm. ⁷

Literature highlighting the use of PCD-CT’s increased spatial resolution in pediatric applications is limited. An early study used a prototype Si-based PCD-CT to evaluate low- and high-contrast tasks on pediatric phantoms demonstrated a higher detectability index for PCD-CT over EID-CT across a range of tube potential and tube current settings. ⁴⁴ The improved detectability index accounts for both the increased spatial frequency and signal-to-noise properties of PCD-CT over EID-CT.

However, there is robust evidence in the adult literature of the superior spatial resolution achieved by PCD-CT compared to EID-CT. This was shown first on phantom studies using prototype PCD systems. ^{8,11,36,45,46} The first musculoskeletal applications performed in animals and human cadavers showed improved visualization of bone microarchitecture on PCD-CT compared to EID-CT, including cortical and trabecular detail and nutrient canals, with substantial reductions in radiation dose and image noise. ^1–3 These advantages were validated in adult patients in studies performed on the wrist, shoulder, and pelvis. ^16,42 PCD-CT has improved the visualization of metastatic bony lesions from breast cancer and multiple myeloma compared to EID-CT. ^47,48 Similar results are expected to be applicable to pediatric oncology as well. Figure 2 illustrates the improved spatial resolution of bony detail on images of the elbow in a pediatric patient. Figure 3 shows a scaphoid fracture that was occult on plain radiographs, but readily visible on PCD-CT.

Open in a separate window

Figure 2.

A 13-year-old male with right elbow pain underwent clinical CT. Both initial EID-CT (a) and follow-up PCD-CT (b) 6 months later show an osteochondral lesion of the capitellum. The follow-up PCD-CT shows healing changes of the OCD (arrow). The lesion is more sharply demarcated on the PCD-CT, which is routinely obtained at thinner slices and sharper kernels (0.6mm, Br89u) than EID-CT (1mm, Ur77u). Despite improved spatial resolution, the PCD images have similar noise and lower dose with similar scanning parameters (PCD: 36 HU, CTDIvol 6.82 mGy vs EID 30 HU, CTDIvol 9.16 mGy) due to smaller detector subpixels and improved dose efficiency. EID, energy integrated detector; HU, Hounsfield unit; PCD, photon counting detector.

Open in a separate window

Figure 3.

A 10-year-old female with a history of falling on the right outstretched hand. No fractures were detected on initial radiographs (not pictured), or follow-up obtained 3 weeks after injury (a), but an MRI was recommended to exclude an occult fracture. MRI was performed a month after the injury showing a faint linear T1 hypointense (b), T2 hyperintense (c) line compatible with a nondisplaced fracture of the scaphoid despite some motion artifact. Subsequently, 3months after injury, the patient had persistent pain and a PCD-CT scan was acquired showing a subtle longitudinal subcortical, non-displaced fracture extending through the lateral cortex (arrow) of the mid-right scaphoid shown on the coronal (d) and (e) image. PCD, photon counting detector.

The spatial resolution advantages of PCD-CT in musculoskeletal imaging are also applicable to temporal bone imaging. Cadaveric and adult studies have documented superior spatial resolution and improved ability to display the ossicular chain and incudomalleolar and incudostapedial joints at substantially lower image noise, and reduced radiation dose, through the elimination of the comb or grid attenuating filter (Figure 4). ^6,49

Open in a separate window

Figure 4.

A 16-year-old male with a history of right cholesteatoma was imaged with PCD-CT to evaluate a right ossicular prosthesis. Images from 3D reconstruction, 0.2mm, kernel Hr84u. The prosthesis articulates with the native stapes (arrow) which is well situated at the oval window (a). However, there is no apparent articulation laterally with the remaining ossicular chain suggesting displacement/dislocation (arrow) in (b). 3D, three-dimensional; PCD, photon counting detector.

The increased spatial resolution of PCD-CT has also been emphasized in high-resolution imaging of the lung, where it has improved visualization of third- and fourth-order bronchi, lung nodules, and increased reader confidence in identifying reticulation, ground-glass opacities, and mosaic pattern in usual interstitial pneumonia, ^50,51 with better lung nodule characterization and volume estimation. ⁵² Figure 5 highlights the increased spatial resolution of PCD compared to EID-CT to identify small peripheral structures in the lung of a pediatric patient with hereditary hemorrhagic telangiectasia syndrome.

Open in a separate window

Figure 5.

An example of multienergy chest CT with intravenous contrast acquired using the ultra-high resolution mode with a PCD-CT in a 12-year-old female with hereditary hemorrhagic telangiectasia. A previous EID-CT 0.75mm image (a) displays only one of three small, peripheral arteriovenous malformations (arrowheads) shown on the PCD-CT 0.8mm image (b) at the same level. In addition to high-resolution images, simultaneous multienergy acquisition on PCD-CT displays the heterogeneous perfusion of the lung parenchyma (c) and permits calculation of lung blood perfusion volume. EID, energy integrated detector; PCD, photon counting detector.

Published pediatric PCD-CT studies of the lung have thus far emphasized radiation dose reduction over maximizing spatial resolution. ^4,29,30 Cao et al have suggested that the increased spatial resolution of PCD-CT may translate to earlier detection of metastatic lung nodules, and the added benefit of increased contrast resolution may be used to better visualize small cardiac structures in pediatric patients. ²⁸

Radiation dose reduction and ultra-low-dose CT

The rapid rise of CT utilization among children in the past decades has prompted concerns regarding the potential long-term effects of ionizing radiation in this patient population. ⁵³ Professional organizations and activities, such as Image Gently and Step Lightly, have helped to educate pediatric imaging providers on the importance of and methods for optimizing radiation doses in children, ^54,55 resulting in a reduction in CT utilization in pediatric patients in the last decade. ^56,57 Meanwhile, doses from CT exams continue to decrease, ⁵⁸ in part due to use of newer dose reduction technologies and requirements to review CT protocols and compare doses against peer organizations. ⁵⁹

The optimization of pediatric CT protocols emphasizes radiation dose reduction and rapid scanning. ^17,18,41 Multiple studies have shown a range of achievable radiation dose reductions in PCD-CT over EID-CT; the amount of reduction depending on the acquisition protocol and diagnostic task. Using a PCD prototype system to scan pediatric phantoms, Chen et al was able to show the equivalent of a 30% dose reduction for a soft tissue detection task and a 70% dose reduction for an iodine detection task. ⁴⁴ In a retrospective review of 27 children who underwent PCD-CT and were compared to an age-matched cohort with the same water-equivalent diameter scanned on EID-CT, Siegel et al showed a lower median volume CT dose index (CTDIvol) of 0.41 vs 0.71 mGy, p < .001, and a lower size-specific dose estimate (SSDE) of 0.82 vs 1.34 mGy, p < .001 ⁴ , on PCD-CT compared to EID-CT. Lung attenuation, noise, and signal-to-noise ratio (SNR), however, were equivalent, and subjective assessment showed no differences in image quality. ⁴ In a study comparing contrast-enhanced chest PCD-CT performed at 90 kV to dual-source EID-CT at 70 kV in children under 3 years with congenital heart disease, PCD-CT showed equivalent radiation dose to the EID-CT scans, while increasing the SNR and CNR. ⁶⁰ Likewise, numerous adult studies have shown that radiation dose may be significantly reduced in PCD-CT with similar or better image quality than EID-CT. Examples include up to a 32% dose reduction in contrast-enhanced abdominopelvic CT for general oncologic imaging, ⁵ a 49% dose reduction for routine CT imaging of the wrist, ¹⁶ and a 54% radiation dose reduction for the detection of focal liver lesions. ⁶¹ The dose reduction in PCD-CT is partially attributable to the elimination of septa from the detector design. Further, with an inherently higher system spatial resolution, reconstruction kernels can incorporate smoothing—which lowers noise—and still achieve the same spatial resolution as an EID-CT protocol at a lower dose. This effect is due to the small detector element size. ^8,14,62–64

A unique quality of PCD-CT that allows radiation dose reduction is its ability to eliminate electronic noise. Quantum noise arises from the stochastic (random) nature of the photon interaction processes, whereas electronic noise arises from the analog electronic circuits within the CT hardware itself. ¹¹ In PCD-CT, a low energy threshold less than 20 or 25 keV can be set to eliminate small signals attributable to electronic noise, resulting in noticeably higher quality images at low signal levels to the detector, which occurs at low radiation doses (Figure 3). ^65,66 Reduced electronic noise is also beneficial in obese patients, or in regions of the body prone to streak artifacts (e.g. the shoulders or pelvis), where attenuation is high. It is particularly beneficial in low-dose imaging, such as screening exams, where there is low incident photon flux and the electronic noise represents a greater proportion of the total signal. ^9,11,67

Further radiation dose reduction can be achieved through “spectral shaping,” a technique that is not novel to PCD-CT. ^68–72 Spectral shaping is most effectively utilized in imaging bone and lung parenchyma, where a tin filter is used to filter out the low energy photons that are mostly absorbed by biologic tissues, but do not contribute appreciably to image signal. ^14,68,73 This technique has been shown in the literature to produce non-contrast chest CT images of diagnostic image quality in pediatric patients, including inspiratory and expiratory phases, down to a mean effective dose of 0.12 ± 0.04mSv, with a CTDIvol of 0.07 ± 0.03mGy per scan. ^29,30 Figure 6 shows an ultra-low dose CT in a patient with cystic fibrosis (CTDIvol of 0.05 mGy); it also shows a radiation dose reduction of about 50% in same-day, EID vs PCD high-resolution chest CTs in another pediatric patient. Additional applications for low-dose imaging are relevant wherever reduced dose imaging is a high priority, such as in patients that require serial CT imaging or have increased sensitivity to radiation. Low-dose PCD-CT imaging with a tin filter has also shown radiation dose reductions compared to EID-CT of up to 85% in the temporal bone CT, ^14,74 67% in sinus CT, ¹⁴ 50% in imaging femoral acetabular impingement, ⁷⁵ and 66% in high-resolution lung imaging in adults. ⁷⁶

Open in a separate window

Figure 6.

A 6-year-old female with a history of cystic fibrosis underwent clinically-indicated PCD-CT as part of routine clinical follow-up. Coronal inspiratory (a) and expiratory (b) PCD-CT images show cylindrical bronchiectasis and air trapping, respectively in the right lower lobe. A cropped image highlights the bronchiectasis in the right lower lobe before (c) and after (d) post-processing with convolutional-neural network denoising. Scan parameters: 100 kV with a tin filter, CTDIvol: 0.05 mGy inspiration and 0.05 mGy expiration. Similarly, a 17-year-old with dyspnea and exercise-induced asthma underwent same day EID- and PCD-CT as part of a research study. (e) The dose of each inspiratory and expiratory PCD scan (CTDIvol 0.21 mGy, effective dose 0.16mSv) is about half the dose of each EID scan shown in (f) (CTDIvol 0.43 mGy, effective dose 0.37mSv). The air-trapping (black arrow) in the left lower lobe is well-demarcated and the mild peribronchial thickening (white arrow) in the central portions of the left upper lobe bronchi are well seen on both scan types. Similar levels of iterative reconstruction were used, with the PCD image showing less noise (PCD-CT 56 HU vs EID-CT 88 HU, with the ROI measured in the ascending aorta). EID, energy-integrated detector; PCD, photon counting detector.

Improved iodine signal

Unlike EID-CT, where the signal recorded is the summation of all photon energies, individual incident photons and their energy information are recorded in PCD-CT. ¹⁰ The photon counting detection mechanism allows for lower energy photons, which carry more iodine signal, to be more optimally weighted in the output signal than in EID-CT, ⁷⁷ thereby increasing iodine SNRs. ¹²

Strategies to increase iodine signal at CT have long incorporated the concept that iodine attenuation increases at lower photon energies, which in EID-CT is typically accomplished using lower tube potentials. ^78–82 Phantom studies demonstrated a mean increase in iodine CNR of 11, 23, 31, and 38% at 80, 100, 120, and 140 kV in PCD-CT over EID-CT. ¹²

With its energy-resolved information, PCD-CT also allows for the reconstruction of virtual monoenergetic images (VMIs) corresponding to lower photon energies, just as is done in dual-energy CT. Studies have documented up to 75% higher CNR for VMIs below 60 keV obtained at 90kV. ⁸³ Multiple studies in adults have suggested 45–50 keV as the optimal VMI range to assess small vessels, the aorta, and the visualization of pulmonary emboli. ^84–88 Low keV VMIs also have been demonstrated to increase conspicuity of hypovascularized hepatic metastases through improved iodine CNR. ⁸⁹ From physics principles, the effect should be even more pronounced in small patients, where noise does not increase as dramatically as VMI energy level is decreased.

Because iodine CNR is improved with PCD-CT compared to EID-CT performed with similar scan parameters, there is better visualization of vessels and the interfaces of enhancing tissues on PCD-CT (Figure 7). Cao et al suggests that increased iodine contrast in PCD-CT also could be used to reduce intravenous contrast doses. ²⁸ A contrast media dose reduction of 25% for PCD-CT angiography of the aorta was shown in adult patients to maintain image quality compared to EID-CT images obtained at equivalent radiation doses. ⁸⁶ Sawall et al. estimated that a reduction in iodinated contrast volume of up to 13–37% is feasible with the greatest reduction seen in larger patient sizes. ⁹⁰

Open in a separate window

Figure 7.

10-year-old male patient with Crohn’s disease. (a) Previous EID-CT scan. (b–e) PCD- CT scan. Coronal MIP images (a, b) at 8mm show the portal vein, the PCD-CT (b) provides better visualization of the mesenteric veins as well. (c) A coronal image outlines mural hyper-enhancement and wall thickening in the J-Pouch compatible with pouchitis (arrows). A 50 keV axial image(d) illustrates how lower-energy VMIs may further optimize contrast enhancement compared to 120 kV images (e). EID-CT and PCD-CT scan parameters: 120 kV, MIP coronal images at 8mm (a, b), coronal and axial at 2mm slice thickness (c-e). CTDIvol 2.6 mGy (EID-CT) vs 1.67mGy (PCD-CT). EID, energy-integrated detector; MIP, maximum intensity projection; PCD, photon counting detector; VMI, virtual monoenergetic image.

Multienergy imaging applications

In addition to VMI, the inherent multienergy information acquired by PCD-CT can be used to perform material-specific imaging, including creating iodine maps, virtual non-contrast images (Figure 8b), virtual non-calcium images (Figure 8c, Figure 9), and kidney stone characterization (Figure 10). These applications are similar to those achieved with dual-energy EID-CT. However, PCD-CT offers the ability to perform these tasks with simultaneous high-pitch, high-temporal resolution scanning (i.e. in the “Flash” mode), or ultra-high spatial resolution imaging (“UHR” mode), which has not been possible previously.

Open in a separate window

Figure 8.

A 16-year-old was diagnosed with a high-grade chondroblastic osteosarcoma. A PCD- CT angiogram was performed and (a) shows severe stenosis of the proximal third of the left subclavian artery by the tumor. Using the VNC in (b), the osseous component of the tumor is identified as separate from the iodine-filled neovessels in the tumor in (a). An example of a VNCA reconstruction is shown in (c). This example showcases PCD-CT multienergy capabilities. EID, energy integrated detector; PCD, photon counting detector; VNC, virtual non-contrast; VNCA, virtual non-calcium.

Open in a separate window

Figure 9.

A 7-year-old female, also with scaphoid fracture not visible on radiograph. An axial (a) PCD-CT image of the left wrist shows cortical disruption. Virtual non-calcium images (b) were reconstructed using a proprietary algorithm that reveal increased attenuation consistent with edema/hemorrhage at the fracture site (arrow). PCD, photon counting detector.

Open in a separate window

Figure 10.

A 16-year-old female patient was evaluated with non-contrast CT for renal stones. There is a left staghorn calculus (a). A proprietary algorithm was used to characterize the staghorn calculus as composed of calcium oxylate (b), similar to the calcium depicted in the adjacent bony vertebral elements.

The combination of inherent spectral capability and high spatial resolution allows for material decomposition of smaller structures with less artifact. PCD-CT has been shown to increase the accuracy of smaller renal stone characterization relative to EID-CT in adults. ⁹¹ In addition, the inherent spectral capability of PCD-CT allows for the combination of novel material-specific imaging algorithms and contrast agents, opening up the possibility of visualizing new features of tumor and disease composition. ^92–94

These multienergy capabilities also can be combined with high-pitch (pitch = 3.2) cardiac (66ms temporal resolution) scanning to simultaneously reduce motion artifact and radiation dose. ^7,95,96 One application is to improve the visualization of cardiac anomalies in pediatric patients (Figure 11).

Open in a separate window

Figure 11.

A 16-year-old male with a suspected coronary anomaly. Cardiac multienergy PCD scan reconstructions at (a) 40 keV, (b) 50 keV, (c) 100 keV, (d) color-coded iodine overlay, (e) iodine overlay and (f) VNC. No coronary anomaly was detected but a short segment myocardial bridge (arrow) of the distal left anterior descending artery at the level of the apex was detected as shown in image (g). PCD, photon counting detector; VNC, virtual non-contrast.

Finally, PCD-CT provides multiple pathways to reducing metallic artifacts (Figure 12), including the use of high energy VMI images, tin filtration, and iterative metal artifact reduction (using either all the incident photons, or only those above a preset energy threshold). ²⁷ Clinical applications include reducing streak artifact for orthopedic hardware, but similar techniques can help eliminate streak artifact from the shoulders or contrast injections.

Open in a separate window

Figure 12.

A 17-year-old female with a history of neurogenic bladder, scoliosis, and secondary xanthogranulomatous pyelonephritis of the right kidney. Clinical conventional EID CT scan demonstrating the known kidney pathology with streaking artifacts from spinal metal hardware (arrows) (a). Figures (b–e) show different PCD-CT reconstructions; 120 kV PCD-CT image, also showing streak artifacts [arrows] (b), with 130 keV (c), IMAR threshold low (d), and threshold high (e) images demonstrating several alternatives at PCD for minimizing metallic artifact like using high energy VMIs, IMAR images, threshold high images (which are reconstructed using high-energy incident photons). CTDI vol for EID-CT scan 3.11 mGy vs 2.49 mGy for PCD-CT. EID, energy integrated detector; IMAR, iterative metal artifact reduction; PCD, photon counting detector; VMI, virtual monoenergetic image.

Conclusion

PCD-CT is suited to address many challenges in pediatric imaging. Its superior spatial resolution benefits pediatric patients with their small anatomic structures. The increased dose efficiency is also desirable, with dramatic dose reduction possible for non-contrast imaging with the use of tin filtration, potentially allowing ultra-low-dose CT that replaces plain film radiography for some diagnostic tasks. Improved iodine contrast and VMI may improve detection of subtle abnormalities in oncology, with high pitch imaging negating the need for sedation in some pediatric CT imaging. Further work is needed by the pediatric radiology community to take full advantage of the benefits of PCD-CT for our patients.

Footnotes

REFERENCES