Towards an AI-Driven Data Reduction Framework for Smart City Applications.

The accelerated development of technologies within the Internet of Things landscape has led to an exponential boost in the volume of heterogeneous data generated by interconnected sensors, particularly in scenarios with multiple data sources as in smart cities. Transferring, processing, and storing a vast amount of sensed data poses significant challenges for Internet of Things systems. In this sense, data reduction techniques based on artificial intelligence have emerged as promising solutions to address these challenges, alleviating the burden on the required storage, bandwidth, and computational resources. This article proposes a framework that exploits the concept of data reduction to decrease the amount of heterogeneous data in certain applications. A machine learning model that predicts a distortion rate and its corresponding reduction rate of the imputed data is also proposed, which uses the predicted values to select, among many reduction techniques, the most suitable approach. To support such a decision, the model also considers the context of the data producer that dictates the class of reduction algorithm that is allowed to be applied to the input stream. The achieved results indicate that the Huffman algorithm performed better considering the reduction of time-series data, with significant potential applications for smart city scenarios.

Mass resolution of linear quadrupole ion traps with round rods.

RATIONALE: Auxiliary dipole excitation is widely used to eject ions from linear radio-frequency quadrupole ion traps for mass analysis. Linear quadrupoles are often constructed with round rod electrodes. The higher multipoles introduced to the electric potential by round rods might be expected to change the ion ejection process. We have therefore investigated the optimum ratio of rod radius, r, to field radius, r0, for excitation and ejection of ions. METHODS: Trajectory calculations are used to determine the excitation contour, S(q), the fraction of ions ejected when trapped at q values close to the ejection (or excitation) q. Initial conditions are randomly selected from Gaussian distributions of the x and y coordinates and a thermal distribution of velocities. The N = 6 (12 pole) and N = 10 (20 pole) multipoles are added to the quadrupole potential. Peak shapes and resolution were calculated for ratios r/r0 from 1.09 to 1.20 with an excitation time of 1000 cycles of the trapping radio-frequency. RESULTS: Ratios r/r0 in the range 1.140 to 1.160 give the highest resolution and peaks with little tailing. Ratios outside this range give lower resolution and peaks with tails on either the low-mass side or the high-mass side of the peaks. This contrasts with the optimum ratio of 1.126-1.130 for a quadrupole mass filter operated conventionally at the tip of the first stability region. With the optimum geometry the resolution is 2.7 times greater than with an ideal quadrupole field. Adding only a 2.0% hexapole field to a quadrupole field increases the resolution by a factor of 1.6 compared with an ideal quadrupole field. Addition of a 2.0% octopole lowers resolution and degrades peak shape. With the optimum value of r/r0 , the resolution increases with the ejection time (measured in cycles of the trapping rf, n) approximately as R0.5 = 6.64n, in contrast to a pure quadrupole field where R0.5 = 1.94n. CONCLUSIONS: Adding weak nonlinear fields to a quadrupole field can improve the resolution with mass-selective ejection of ions by up to a factor of 2.7. The optimum ratio r/r0 is 1.14 to 1.16, which differs from the optimum ratio for a mass filter of 1.128-1.130.

Mass selectivity of dipolar resonant excitation in a linear quadrupole ion trap.

RATIONALE: For mass analysis, linear quadrupole ion traps operate with dipolar excitation of ions for either axial or radial ejection. There have been comparatively few computer simulations of this process. We introduce a new concept, the excitation contour, S(q), the fraction of the excited ions that reach the trap electrodes when trapped at q values near that corresponding to the excitation frequency. METHODS: Ion trajectory calculations are used to calculate S(q). Ions are given Gaussian distributions of initial positions in x and y, and thermal initial velocity distributions. To model gas damping, a drag force is added to the equations of motion. The effects of the initial conditions, ejection Mathieu parameter q, scan speed, excitation voltage and collisional damping, are modeled. RESULTS: We find that, with no buffer gas, the mass resolution is mostly determined by the excitation time and is given by R~dß/dq qn, where ß(q) determines the oscillation frequency, and n is the number of cycles of the trapping radio frequency during the excitation or ejection time. The highest resolution at a given scan speed is reached with the lowest excitation amplitude that gives ejection. The addition of a buffer gas can increase the mass resolution. The simulation results are in broad agreement with experiments. CONCLUSIONS: The excitation contour, S(q), introduced here, is a useful tool for studying the ejection process. The excitation strength, excitation time and buffer gas pressure interact in a complex way but, when set properly, a mass resolution R0.5 of at least 10,000 can be obtained at a mass-to-charge ratio of 609.

Optimization of PACS data persistency using indexed hierarchical data.

We present a new approach for the development of a data persistency layer for a Digital Imaging and Communications in Medicine (DICOM)-compliant Picture Archiving and Communications Systems employing a hierarchical database. Our approach makes use of the HDF5 hierarchical data storage standard for scientific data and overcomes limitations of hierarchical databases employing inverted indexing for secondary key management and for efficient and flexible access to data through secondary keys. This inverted indexing is achieved through a general purpose document indexing tool called Lucene. This approach was implemented and tested using real-world data against a traditional solution employing a relational database, in various store, search, and retrieval experiments performed repeatedly with different sizes of DICOM datasets. Results show that our approach outperforms the traditional solution on most of the situations, being more than 600 % faster in some cases.

Trajectory calculations of space-charge-induced mass shifts in a linear quadrupole ion trap.

RATIONALE: If too many ions are stored in a linear quadrupole ion trap, space charge causes the oscillation frequencies to decrease. Ions therefore appear at higher apparent mass-to-charge ratios in a mass spectrum. To further understand this process, we have used trajectory calculations of ions to determine mass shifts. METHODS: Two models of the ion cloud are used. The first assumes that the acceptance of the quadrupole is uniformly filled with ions. The second assumes that the ions have a thermal distribution trapped in the effective potential. Both give analytical descriptions of the field from space charge. Ion trajectories are calculated with and without space charge. Oscillation frequencies are determined with a Fourier transform. Shifts in oscillation frequency with space charge are then used to calculate mass shifts. RESULTS: Both ion cloud models give similar results. More diffuse ion clouds or ion clouds that have higher temperatures produce lower electric fields near the center of the trap and hence lower mass shifts. Space charge produces a nonlinear field. As a result, the discrete resonant frequencies of ions in a pure quadrupole field become distributions of frequencies. Comparisons with experiments show agreement for reasonable values of the parameters of the two ion cloud models. CONCLUSIONS: This relatively simple method for calculating the effects of space charge shows (i) that the spread of oscillation frequencies reduces mass resolution with axial ejection and (ii) that mass shifts are reduced with ion clouds with greater spatial extents or higher ion temperatures.

Space-charge effects with mass-selective axial ejection from a linear quadrupole ion trap.

Methods to reduce mass shifts caused by space charge with mass-selective axial ejection from a linear quadrupole ion trap are investigated. For axial ejection, dipole excitation is applied to excite ions at q ≈ 0.85. The trapping radiofrequency (rf) voltage is scanned to bring ions of different m/z values into resonance for excitation. In the fringing field at the quadrupole exit, excited ions gain axial kinetic energy, overcoming the trapping potential, and are ejected from the trap. Space charge causes the frequencies of ion oscillation to decrease. Thus, greater rf voltages are required to bring ions into resonance for excitation and ejection, and the ions shift to higher apparent masses in a mass spectrum. At the same time, the peaks broaden, lowering resolution. The effects of injection q value, ejection q value, excitation amplitude, quadrupole dc voltages applied to the electrodes, applying an rf voltage to the exit lens, and scan speed, on mass shifts have been studied experimentally. Most experiments were done with only ions of protonated reserpine (m/z 609.3 and its isotopic peaks) in the trap. Some experiments were done with ions of protonated reserpine and ions of m/z 622 in the trap. In general, the mass shifts are reduced with higher ejection q values, higher excitation amplitudes, with quadrupole dc applied, and at higher scan speeds. The application of quadrupole dc appears to increase the ion cloud temperature, which lowers mass shifts. Thus, a proper choice of operating conditions can reduce, but not eliminate, mass shifts caused by space charge.

Linear quadrupoles in mass spectrometry.

The use of linear quadrupoles in mass spectrometry as mass filters and ion guides is reviewed. Following a tutorial review of the principles of mass filter operation, methods of mass analysis are reviewed. Discussed are extensions of quadrupole mass filters to higher masses, scanning with frequency sweeps of the quadrupole waveform, operation in higher stability regions, and operation with rectangular or other periodic waveforms. Two relatively new methods of mass analysis the use of "islands of stability" and "mass selective axial ejection" are then reviewed. The optimal electrode geometry for a quadrupole mass filter constructed with round rods is discussed. The use of collisional cooling in quadrupole ion guides is discussed along with ion guides that have axial fields. Finally, mass analysis with quadrupoles that have large distortions to the geometry and fields is discussed. An Appendix gives a brief tutorial review of definitions of electrical potentials and fields, as well as the units used in this article.

Quadrupole mass filters with added hexapole fields.

Conventional mass analysis has been investigated experimentally with six quadrupole mass filters with added hexapole fields; three with added hexapole fields of 4%, 8% and 12% with equal diameter rods, and three with added hexapole fields of 4%, 8% and 12% with unequal diameter rods to remove an added octopole field. Compared with conventional quadrupoles, these rod sets have very large field distortions. With the positive resolving dc applied to the y rods (Mathieu parameter a(x) < 0) only low resolution (10-100) and low transmission are seen. With the polarity reversed (a(x) > 0) much higher resolution (> or = 1000) and transmission are possible. Increasing the magnitude of the added hexapole field decreases the limiting resolution at m/z 609. Removing the added octopole field increases the limiting resolution. In some cases structure is formed on the peaks. For a given scan line slope, U/V(rf), the resolution decreases as the amplitude of the added hexapole field increases. These results are consistent with changes to the stability diagrams, calculated here. With a(x) > 0, adding a hexapole field causes the x stability boundary to move outward with all rod sets. With a(x) < 0, the boundaries become diffuse and the tip of the stability diagram becomes rounded, limiting the resolution to ca. 10-100. Where comparisons are possible, experiments show the rod sets with added hexapole fields have transmission 10-300 times less than a conventional quadrupole. Thus these quadrupoles are less useful for mass analysis than conventional quadrupoles. However, it is surprising, given the highly distorted fields, that some of the quadrupoles give resolution of 1000 or more.

Overcoming field imperfections of quadrupole mass filters with mass analysis in islands of stability.

We have constructed, and tested as mass filters, linear quadrupoles with added hexapole fields of 4%, 8%, and 12%, with and without added octopole fields. A hexapole field can be added to the field of a linear quadrupole by rotating the two y rods toward an x rod. This also adds an octopole field which can be removed by making the x rods greater in diameter than the y rods. In comparison to conventional quadrupole mass filters these rod sets have severely distorted quadrupole fields, with a mix of both even and odd higher spatial harmonics. They allow evaluating the performance of rod sets with strong geometric and field distortions as mass filters. Conventional mass analysis at the tip of the stability diagram has been compared to mass analysis using islands of stability. The stability islands are produced by applying an auxiliary quadrupole excitation field to the quadrupole. We show that with normal mass analysis at the tip of the stability diagram, the transmission, resolution, and peak shapes are relatively poor in comparison to a conventional rod set. However, the use of islands of stability dramatically improves the resolution and peak shape, and in some cases ion transmission, suggesting that mass analysis with islands of stability may provide a method to overcome a wide range of field imperfections in linear quadrupole mass filters.

Mass selective axial ion ejection from linear quadrupoles with added octopole fields.

Mass selective axial ejection of ions from linear quadrupoles with added octopole fields is described. Quadrupoles with 2.0% and 2.6% added octopole fields have been tested and compared with a conventional quadrupole. The effects of trapping ions at different q values, excitation voltage, scan direction, balanced and unbalanced rf voltages on the rods, and dc applied between the rods have been investigated. The highest scan speeds and best resolution are obtained with resonant excitation and ejection at high q (q = 0.8). With axial ejection, the quadrupole with a 2.0% added octopole field provides mass resolution and ejection efficiencies similar to a conventional rod set. Quadrupole, dipole, and simultaneous dipole-dipole excitation between the x and y rod pairs were compared, and no advantage was found with quadrupole or dipole-dipole excitation. The effects of scan speed were investigated and a resolution at half height of about 1600 is possible at scans speed up to 5000 Th/s.

Quadrupole excitation of ions in linear quadrupole ion traps with added octopole fields.

Modeling and experimental studies of quadrupole excitation of ions in linear quadrupole traps with added octopole fields are described. An approximate solution to the equations of motion of ions trapped in a quadrupole with added octopole and dodecapole fields, with quadrupole excitation and damping is given. The solutions give the steady-state or stationary amplitudes of oscillation with different excitation frequencies. Trajectory calculations of the oscillation amplitudes are also presented. The calculations show that there can be large changes in the amplitude of ion oscillation with small changes in excitation frequency, on both the low and high-frequency sides of a resonance. Results of experiments with quadrupole excitation of reserpine ions in linear quadrupole traps with 2.0%, 2.6%, and 4.0% added octopole fields are given. It is found that as the excitation frequency is changed, two resonances are generally observed, which are attributed to the motion in the x and y directions. The two resonances can have quite different intensities. Sudden jumps or sharp sided resonances are not observed, although in some cases asymmetric resonances are seen. The calculated frequency differences between the two resonances are in approximate agreement with the experiments.

Conformations of gas-phase ions of ubiquitin, cytochrome c, apomyoglobin, and beta-lactoglobulin produced from two different solution conformations.

At low pH in solutions of 50% methanol, proteins form expanded denatured states (the "H" state). In 90% methanol, proteins form expanded helical denatured states with artificial alpha-helices (the "H(c)" state). Gas-phase ions of ubiquitin, cytochrome c, apomyoglobin, and native and disulfide-reduced beta-lactoglobulin were formed by electrospray ionization (ESI) of the proteins from the H and H(c) states in solution. Both states in solution produce the same charge states in ESI. The conformations of the ions were studied with cross section measurements and gas-phase H/D exchange experiments. The cross sections show that the ions retain considerable folded structure. For a given protein and given charge state, ions produced from the H and H(c) states showed the same cross sections (within approximately 1%). Ions of cytochrome c, apomyoglobin, and native and reduced beta-lactoglobulin of a given charge state showed no differences in H/D exchange level when produced from the H or H(c) state. However, ubiquitin ions produced from the H(c) state consistently exchange fewer ( approximately 13%) hydrogens than ions produced from the H state, suggesting that in this case the gas-phase protein ions retain some memory of their solution conformations.

Mass analysis in islands of stability with linear quadrupoles with added octopole fields.

Mass analysis with linear quadrupole mass filters is possible by forming "islands" in the stability diagram with auxiliary quadrupole excitation. In this work, computer simulations are used to calculate stability boundaries, island positions, and peak shapes and ion transmission for mass analysis with linear quadrupole mass filters that have added octopole fields of about 2 to 4%. Rod sets with exact geometries that have quadrupole and octopole fields only in the potential, and round rod sets, with multipoles up to N = 10 (the twenty pole term) included in the calculations, show the same stability boundaries, island positions, and peak shapes. With the DC voltage applied to the rods so that the Mathieu parameter a < 0, conventional mass analysis is possible without the use of an island. With the DC polarity reversed so that a > 0, the resolution and transmission are poor preventing conventional mass analysis. In principle, mass analysis in an island is possible with operation at either of two tips. Provided the correct island tip is chosen for mass analysis, peak shapes comparable to those with a > 0 and no excitation are possible, both with a > 0 and with a < 0. In the latter case, the use of an island of stability allows mass analysis when the added octopole otherwise prevents conventional mass analysis.

Linear quadrupoles with added hexapole fields.

Linear quadrupoles with added hexapole fields are described. The shifts in ion oscillation frequency caused by the addition of a hexapole field are calculated within the effective potential model. Methods to construct linear quadrupoles with added hexapole fields with exact electrode geometries and with round rods are discussed. A quadrupole with added hexapole field can be constructed with round rods by rotating two rods (say the y rods) towards an x rod. Computer simulations are used to investigate the possibility of mass analysis with quadrupoles with added hexapole fields. We find that a quadrupole with an added hexapole field in the range 2-12% can provide mass analysis provided the dc is applied with the correct polarity and value. When a rod set is constructed with round rods, other multipoles in the potential degrade the peak shape, resolution and transmission. The largest of these after the quadrupole and hexapole are a dipole and octopole term. With round rod sets, the peak shape can be improved by using different diameters for the x and y rod pairs to minimize the octopole term in the potential and by injecting ions at the field center where the dipole term is zero. Calculations of the boundaries of the stability diagram for this case show the boundaries move out, relative to those of a pure quadrupole field, but remain sharp.

Ion excitation in a linear quadrupole ion trap with an added octopole field.

Modeling of ion motion and experimental investigations of ion excitation in a linear quadrupole trap with a 4% added octopole field are described. The results are compared with those obtained with a conventional round rod set. Motion in the effective potential of the rod set can explain many of the observed phenomena. The frequencies of ion oscillation in the x and y directions shift with amplitude in opposite directions as the amplitudes of oscillation increase. Excitation profiles for ion fragmentation become asymmetric and in some cases show bistable behavior where the amplitude of oscillation suddenly jumps between high and low values with very small changes in excitation frequency. Experiments show these effects. Ions are injected into a linear trap, stored, isolated, excited for MS/MS, and then mass analyzed in a time-of-flight mass analyzer. Frequency shifts between the x and y motions are observed, and in some cases asymmetric excitation profiles and bistable behavior are observed. Higher MS/MS efficiencies are expected when an octopole field is added. MS/MS efficiencies (N(2) collision gas) have been measured for a conventional quadrupole rod set and a linear ion trap with a 4% added octopole field. Efficiencies are chemical compound dependent, but when an octopole field is added, efficiencies can be substantially higher than with a conventional rod set, particularly at pressures of 1.4 x 10(-4) torr or less.

An aerodynamic drag model for protein ions.

The energy losses of protein ions passing through a collision cell filled with inert gas have been modeled as the aerodynamic drag on a projectile at high Knudsen number. When applied to the energy loss data of Covey and Douglas (J. Am. Soc. Mass Spectrom. 1993, 4, 616-623) with drag coefficients from the gas dynamics literature, derived protein cross sections are ∼0.8 of those found with the simple collision model used by Covey and Douglas.

Collision cross sections for protein ions.

A method for the determination of cross sections for gas-phase protein ions, based on the energy loss of ions as they pass through a collision gas, is described. A simple model relates the energy loss to the number of collisions and hence the cross section. Results from a Monte Carlo model that support the validity of this approach are described. Experimental cross sections are reported for motilin, ubiquitin, cytochrome c, myoglobm, and bovine serum albumin. Cross sections range from approximately 800 Å(2) for motilin to approximately 14,000 Å(2) for bovine serum albumin and generally increase with the number of charges on the ion. Cytochrome c ions from aqueous solution show somewhat smaller cross sections than ions formed from solutions of higher organic content, suggesting that the gas-phase ions may retain some memory of their solution conformation.

Collisional focusing effects in radio frequency quadrupoles.

The transmission of ions through a conventional two-dimensional radiofrequency-only (rf) quadrupole has been studied for comparatively high operating pressures between 5 × 10(-4) and 1 × 10(-2) torr. Measurements of signals from mass-resolved analyte ions and total ion currents show that, provided the initial injection ion energy is low (1-30 eV), the ion transmission observed through a small aperture at the exit of the rf quadrupole first increases as the gas pressure increases, reaching a maximum at - 8 × 10(-3) torr before decreasing at higher pressures. This is in direct contrast to the expectations of classical scattering. This "collisional focusing" appears to be analogous to effects seen in three-dimensional ion traps. The collisional focusing increases with the mass of the ion (not mass-to-charge ratio) for masses up to at least 16,950 u. The collisional focusing of the ions is found to be accompanied by significant losses of axial kinetic energy. A Monte Carlo simulation of the energy loss process is reported that can provide agreement with the observed losses for reasonable collision cross-sections. The results suggest that operation of rf quadrupoles at relatively high pressure may find practical application in sampling ions from high (e.g., atmospheric) pressure ion sources.

H/D exchange of gas phase bradykinin ions in a linear quadrupole ion trap.

The gas phase H/D exchange reaction of bradykinin ions, as well as fragment ions of bradykinin generated through collisions in an orifice skimmer region, have been studied with a linear quadrupole ion trap (LIT) reflectron time-of-flight (rTOF) mass spectrometer system. The reaction in the trap takes only tens of seconds at a pressure of few mTorr of D2O or CD3OD. The exchange rate and hydrogen exchange level are not sensitive to the trapping q value over a broad range, provided q is not close to the stability boundary (q = 0.908). The relative rates and hydrogen exchange levels of protonated and sodiated +1 and +2 ions are similar to those observed previously by others with a Fourier transform ion cyclotron resonance (FTICR) mass spectrometer system. The doubly and triply protonated ions show multimodal isotopic distributions, suggesting the presence of several different conformations. The y fragment ions show greater exchange rates and levels than a or b ions, and when water or ammonia is lost from the fragment ions, no exchange is observed.

Unfolding of proteins monitored by electrospray ionization mass spectrometry: a comparison of positive and negative ion modes.

Electrospray ionization (ESI) mass spectrometry (MS) in both the positive and negative ion mode has been used to study protein unfolding transitions of lysozyme, cytochrome c (cyt c), and ubiquitin in solution. As expected, ESI of unfolded lysozyme leads to the formation of substantially higher charge states than the tightly folded protein in both modes of operation. Surprisingly, the acid-induced unfolding of cyt c as well as the acid and the base-induced unfolding of ubiquitin show different behavior: In these three cases protein unfolding only leads to marginal changes in the negative ion charge state distributions, whereas in the positive ion mode pronounced shifts to higher charge states are observed. This shows that ESI MS in the negative ion mode as a method for probing conformational changes of proteins in solution should be treated with caution. The data presented in this work provide further evidence that the conformation of a protein in solution not its charge state is the predominant factor for determining the ESI charge state distribution in the positive ion mode. Furthermore, these data support the hypothesis of a recent study (Konermann and Douglas, Biochemistry 1997, 36, 12,296-12,302) which suggested that ESI in the positive ion mode is not sensitive to changes in the secondary structure of proteins but only to changes in the tertiary structure.