Page 64 - AIH-2-1

Artificial Intelligence in Health Algorithm and metal oxide nanoparticle in MRI

and (004). An additional peak at 35.58º corresponded to the Figure 4 presents a series of curves demonstrating
(111) plane of the Cu O phase. These peaks were consistent the relationship between signal intensity and NP
2
with the monoclinic phase of CuO (JCPDS: 89-5895), with concentrations, as well as the dependency of signal
lattice parameters a = 4.682 Å, b = 3.424 Å, and c = 5.127 intensity on TE for these NPs in MRI. The results revealed
Å. The FTIR spectrum of these NPs in Figure 2D presents that the pixel signal intensity is inversely proportional to
55
a band at 474 cm , denoting Cu‒O formation. 56,57 the ET and NP concentration, a trend that was consistent
−1
Figure 2E presents the XRD pattern of the Fe O across all the tested NPs. As the ET increased, a reduction
3
2
NPs, with peaks corresponding to rhombohedral Fe O in signal intensity was observed for all NPs. Furthermore,
this reduction in signal intensity was more pronounced
3
2
(JCPDS: 79-0007) and lattice parameters a = b = 5.0285 for NP concentrations with greater metal concentrations.
Å and c = 13.7360 Å. 39,58,59 The FTIR spectrum of these
NPs in Figure 2F reveals absorption bands at 520 cm and Moreover, a steeper slope in the T2 relaxation curve
−1
437 cm , attributed to the Fe‒O stretching and bending correlated with a greater decrease in the signal intensity,
−1
enhancing the effectiveness of the NPs as T2 negative CAs.
modes in hematite. 60
In addition, Figure 4A-E illustrates variations in
In the XRD spectrum of NiO NPs displayed in
Figure 2G, the peaks at 37.50°, 43.50°, 63.05°, and 75.58° the signal intensities for the T1, T2, and FLAIR scan
corresponded to the lattice planes (111), (200), (220), sequences as a function of the metal NP concentration.
All metal oxide NPs exhibited a decrease in the signal
and (311) of the cubic phase of NiO (JCPDS: 01-1239), intensity with increasing NP concentration across all
with lattice parameters a = b = c = 4.1710 Å. 40,61 The FTIR three sequences consistently. Notably, Fe O NPs exhibited
spectrum of these NPs in Figure 2H presents a prominent a more pronounced signal reduction, while ZnO NPs
2
3
absorption band at 669 cm , signifying Ni‒O formation. 62
−1
demonstrated a slightly divergent behavior in the FLAIR
In the XRD spectrum of ZnO NPs displayed in Figure 2I, sequence. Consequently, a steeper relaxation curve slope
the diffraction peaks aligned with the lattice planes (100), correlated with increased signal reduction, enhancing the
(002), (101), (102), (110), (103), (200), (112), (201), (004), efficacy of metal oxide NPs as CAs.
and (202) of hexagonal ZnO (JCPDS: 65-3411), with lattice The magnitude of the T2 contrast effect was
parameters a = b = 3.249 Å and c = 5.206 Å. 63,64 The FTIR quantitatively represented by spin‒spin relaxivity R ; an
spectrum of these NPs in Figure 2J illustrates a band at increase in the R values indicated a corresponding increase
2
−1
2
408 cm , corresponding to Zn‒O bonding, and a band at in the contrast effect. The relaxation rate R (=1/T2) is
669 cm , ascribed to C‒H stretching in the alkyne group. 65,66 plotted against the ET in Figure 5A. An analysis of these
2
−1
The average crystallite size was estimated using the curves revealed that NP concentrations directly impact the
Debye–Scherrer equation relaxivity time, thereby influencing the pixel intensity. In
Figure 5B-D, which corresponds to the T1, T2, and FLAIR
d = 0.89λ/βcosθ, (III)
MRI sequences, respectively, the pixel intensity trends for
where 0.89 is the Debye constant, λ represents the each metal oxide NP were distinctly observed in relation
X-ray wavelength (1.5406 Å), β denotes the full-width at to their concentrations. In particular, Fe O NPs (depicted
3
2
half-maximum of the peak, and θ represents the Bragg by the blue curve) demonstrated the most significant
angle. The estimated average crystallite sizes for Co O , intensity variation across all sequences, showing a notable
4
3
CuO, Fe O , NiO, and ZnO NPs were 12, 26, 21, 38, and intensity decrease with increasing concentrations, which
3
2
25 nm, respectively. highlights their substantial impact on MRI imaging.
Conversely, Co O NPs (black curve) exhibited a more
3.2. Manual evaluation: MRI signal and relaxation gradual decline, suggesting a less pronounced impact of
3
4
time evaluation their concentrations on MRI signal intensities. Other metal
The acquired MRIs were manually evaluated to quantify the oxide NPs, including CuO (red curve), NiO (green curve),
mean signal intensity in each compartment of the phantom and ZnO (purple curve), also exhibited reductions in the
across varying NP concentrations. Figure 3 illustrates the pixel intensities with increasing concentrations; however,
signal intensity as a function of the NP concentration these changes vary, reflecting the distinct reactivity of each
for the T1, T2, and FLAIR sequences. Circles in Figure 3 metal oxide NP in the MRI sequences. These differences
represent the cross-sections of each compartment from underscore the importance of considering the unique
which the mean pixel intensity was extracted. Notably, properties of different metal oxide NPs when employing
different NPs altered the MRI signal characteristics, with them as CAs in MRI to optimize image acquisition and
Fe O NPs exerting particularly notable effects. interpretation.
3
2
Volume 2 Issue 1 (2025) 58 doi: 10.36922/aih.3947