Method and system for improving resolution of a mass spectrometer based on data analysis

Abstract

The application discloses a mass spectrometer resolution improvement method and system based on data analysis, and relates to the technical field of precision detection, and comprises the following steps: S100, in the continuous sampling signal, the phase change of the high-energy section is tracked point by point with fine time granularity, each time phase rise and phase drop are connected in time sequence, and a continuous extension phase fluctuation time band is generated, which is used for constructing a through phase tracking foundation; S200, the adjacent sampling beat difference value is compared around the phase fluctuation time band, the position of the phase chain break is gathered into a list of candidate pieces of sudden jumps in time sequence, and the start point, end point and energy sudden increase position are recorded in each list of candidate pieces of sudden jumps, which are used for subsequent crack positioning. The application realizes abnormal signal repair and peak shape remodeling by constructing a phase fluctuation time band, extracting a sudden jump piece and generating a crack fingerprint, combining fine-grained sampling regulation and spiral rearrangement type phase domestication, improves mass spectrometer resolution and stability, and enhances quantitative accuracy.

Description

Technical Field

[0001] This invention relates to the field of precision detection technology, and more specifically to a method and system for improving the resolution of mass spectrometers based on data analysis. Background Technology

[0002] Data-driven mass spectrometry resolution enhancement refers to the use of in-depth analysis and processing of mass spectrometry signals. Without altering the mass spectrometer's hardware structure, this involves multi-dimensional data processing operations such as noise identification, peak morphology analysis, inter-peak interference elimination, and trajectory reconstruction of the acquired raw mass spectrometry data. This enhances the separation and detail recognition capabilities of mass spectrometry peaks, making adjacent mass numbers easier to distinguish. The core idea is to extend the upper limit of traditional hardware resolution into the optimizable space of software. By precisely reconstructing the statistical characteristics, time-frequency properties, and energy distribution of the mass spectrometry data, interference factors affecting peak resolution, such as peak broadening, drift, and tailing, are eliminated. This results in substantial improvements in peak clarity, inter-peak distance recognition, and the accuracy of weak signal extraction in the final output mass spectrum, thereby enhancing the overall resolution of the mass spectrometer.

[0003] The existing technology has the following shortcomings:

[0004] In existing technologies, mass spectrometers typically rely on increasing the sampling rate to capture subtle changes during ion flight in order to improve resolution. However, in the rapid sampling region, the signal phase is highly susceptible to transient perturbations in the high-energy segment, especially under extreme conditions such as sudden increases in ion density, short-term saturation of the detector link, and instantaneous reverse response of the electron multiplication chain, which can cause extremely short phase jumps. These phase jumps do not manifest continuously but rather disrupt the original phase evolution trajectory as transient breaks, causing discontinuities in the signal waveform of the high-energy segment in the time series, thus forming a broken band in the peak structure.

[0005] Once a break zone appears, the peak shape of that segment will lose its proper symmetry, continuity, and fit, exhibiting irregular oscillations, making it impossible to effectively extract key values ​​such as peak position, peak width, and peak area. This anomaly not only disrupts the overall continuity of the spectrum along the time and mass axes but also causes the baseline extraction chain, peak shape reconstruction chain, and mass calibration chain to all fail in this segment, rendering quantitative models unable to perform intensity integration, mass inversion, and feature identification. More seriously, in continuous scanning mode, this anomaly will gradually amplify along the scanning path, causing multiple adjacent sampling windows to lose stable correlation, resulting in a chain-like paralysis of the entire continuous scanning mode, ultimately rendering the entire round of mass spectrometry analysis completely unusable.

[0006] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for improving the resolution of mass spectrometers based on data analysis, so as to solve the problems in the background art mentioned above.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a mass spectrometer resolution improvement method based on data analysis, comprising the following steps:

[0009] S100 tracks high-energy phase changes point by point in continuously sampled signals with fine time granularity, and connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation.

[0010] S200 compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a list of jump candidate segments in chronological order. The start point, end point and energy surge position are recorded in each jump candidate segment list for subsequent crack location.

[0011] S300 maps each jump fragment to both the time axis and the mass axis based on the list of jump candidate fragments, generating a phase crack fingerprint in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference.

[0012] S400 sets fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map expansion range. Within the adjustment band, the sampling interval and detector gain are fine-tuned so that the phase at both ends of the crack gradually enters the splicable range, forming a time buffer band covering the crack boundary.

[0013] The S500 performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped step by step in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

[0014] Preferably, step S100 includes:

[0015] During the signal sampling stage, the continuous sampling signal is sampled with fine time granularity. The amplitude difference value of each point is calculated and the absolute timestamp is recorded to form the phase rising segment and the phase falling segment.

[0016] After the phase rising segment and the phase falling segment are formed, the phase rising segment and the phase falling segment are connected in chronological order and an amplitude stabilization region is set to form a continuous phase evolution chain.

[0017] After forming a continuous phase evolution chain, the continuous phase path is labeled with intervals according to the duration and amplitude range to filter out stable phase fluctuation intervals.

[0018] After selecting stable phase fluctuation intervals, the original amplitude, derivative value and local extreme state are recorded using the time marker of the stable phase fluctuation interval as a reference to form a set of phase fluctuation time bands.

[0019] Preferably, step S200 includes:

[0020] In the phase fluctuation time band, the phase change amplitude between adjacent sampling beats is calculated point by point according to the sampling time sequence, and the amplitude difference is recorded to form a sequence of adjacent beat difference values.

[0021] After forming the difference sequence of adjacent beats, micro-region analysis is performed on the amplitude abrupt increase points in the difference sequence of adjacent beats in chronological order to identify abrupt jump segments with temporal continuity and significant jump amplitude;

[0022] After identifying the jump segments, the jump segments are numbered in chronological order and the start time, end time, jump duration, and the location of the point with the maximum jump amplitude are recorded to form a list of candidate jump segments.

[0023] After forming a list of candidate jump segments, the list of candidate jump segments is associated with the phase fluctuation time zone, and the original amplitude sequence of the jump segments within the time span is recorded to form a data foundation with complete traceability capability.

[0024] Preferably, after forming a list of candidate jump segments, the amplitude change trend of the sampling points before and after the jump segment is checked by using the start time and end time recorded in the list of candidate jump segments as the boundary. After the continuity check is completed, the amplitude gradient change inside the jump segment is analyzed with the position of the point with the maximum jump amplitude as the center, so as to confirm the abnormal intensity range of the jump segment in the phase fluctuation time band and use it for key weight marking in subsequent biaxial mapping.

[0025] Preferably, step S300 includes:

[0026] Based on the start and end times of each segment recorded in the list of candidate segments for sudden jumps, the phase jump interval is located on the time axis, and the original signal intensity and amplitude trend changes within the phase jump interval are extracted to form time dimension anomalous features.

[0027] After forming time-dimensional anomalous features, the jump segments are mapped to the mass axis interval according to the time marker, and the number of spectral peaks, peak spacing, peak height ratio and signal-to-noise ratio within the mass interval are extracted to form quality-dimensional anomalous features.

[0028] After forming the quality dimension anomaly features, the time dimension anomaly features are fused with the quality dimension anomaly features to depict crack structure stripes with range, direction and intensity in two-dimensional coordinate space;

[0029] After the crack structure strips are formed, they are embedded into the spectral spatial coordinate system in chronological order, and the starting coordinates, coverage area and fingerprint intensity score are recorded to form a phase crack fingerprint map.

[0030] Preferably, after forming the phase crack fingerprint map, the temporal overlap rate and mass overlap rate of adjacent crack structure strips in the phase crack fingerprint map are compared, and when the overlap rate exceeds a preset overlap threshold, the adjacent crack structure strips are merged into crack clusters to form a continuous abnormal region; after forming the continuous abnormal region, the time start point, time end point, mass span and fingerprint intensity of the continuous abnormal region are recorded to improve the crack location accuracy.

[0031] Preferably, step S400 includes:

[0032] Based on the start and end points of the time axis of the crack strips in the phase crack fingerprint image, an extension range is set, and potential sampling adjustment points are divided according to the extension range to form an adjustment interval;

[0033] After the adjustment interval is formed, the sampling interval is finely adjusted point by point according to the distance between each sampling point and the crack center so that the phase change forms a transition gradient on the time axis;

[0034] After forming the transition gradient, the detector gain is fine-tuned at each sampling point within the adjustment interval to enhance the effective identification capability of crack edge signals and form a detection transition zone.

[0035] After forming the detection transition band, the time intervals processed by sampling interval fine-tuning and detector gain fine-tuning are integrated into a continuous buffer band structure to form a stable transition between the crack region and the continuous phase path.

[0036] Preferably, within the adjustment range, the sampling time and amplitude response of the sampling points are synchronously calibrated with the crack center as the reference so that the sampling points inside the buffer zone are distributed in a progressive and continuous manner; after synchronous calibration, the sampling points inside the buffer zone are locally smoothed according to the changing trend of the amplitude response so that the phases at both ends of the buffer zone are gradually aligned and a continuous peak transition path is formed.

[0037] Preferably, step S500 includes:

[0038] The starting center point and path expansion layer of the spiral rearrangement are determined based on the crack length recorded inside the temporal buffer band, and the outer expansion boundary is delineated to form the spiral rearrangement skeleton.

[0039] After forming the spiral rearrangement skeleton, each grid node is assigned a micro offset instruction sequence with hierarchical increments along the spiral path so that the node has time direction offset and mass direction offset;

[0040] After forming the micro-offset instruction sequence, perform micro-displacement operations point by point to rearrange the sampling points in the crack area into the spiral path and record the new sampling timestamp and quality number label;

[0041] After forming the spiral path, full-segment smoothing is performed on the repositioned data points along the spiral path to gradually restore the continuous structure of the peak shape on the time axis and the quality axis.

[0042] A mass spectrometer resolution enhancement system based on data analysis includes a phase fluctuation construction module, a jump fragment extraction module, a crack fingerprint generation module, a temporal buffer band construction module, and a helical rearrangement training module.

[0043] The phase fluctuation construction module tracks the phase changes of the high-energy band point by point in the continuously sampled signal with fine time granularity. It connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation.

[0044] The jump segment extraction module compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a jump candidate segment list in chronological order. The starting point, ending point and energy surge position are recorded in each jump candidate segment list for subsequent crack location.

[0045] The crack fingerprint generation module maps each jump fragment to the time axis and mass axis simultaneously based on the list of jump candidate fragments, generating a phase crack fingerprint map in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference.

[0046] The timing buffer band construction module sets up fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map extension range. Within the adjustment band, the sampling interval fine-tuning and the detector gain fine-tuning are performed so that the phase at both ends of the crack gradually enters the splicable interval, forming a timing buffer band covering the crack boundary.

[0047] The spiral rearrangement training module performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped step by step in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

[0048] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0049] This invention establishes a continuous phase tracking foundation by constructing a phase fluctuation time band, then accurately extracts abrupt jumps and maps them to a two-dimensional space of time and mass axes, forming a crack fingerprint map with directional and spatial coverage characteristics. Combined with fine-grained sampling control and a spiral rearrangement phase conditioning process, dynamic repair and peak reshaping of anomalous signal regions are achieved. Ultimately, in continuous scanning mode, the integrability of peaks and the structural integrity of the spectrum are restored, significantly improving the resolution, stability, and quantitative accuracy of mass spectrometry analysis, expanding the performance ceiling of traditional mass spectrometers, and providing a highly reliable hardware and software integrated enhancement method for high-precision mass spectrometry detection tasks. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0051] Figure 1 This is a flowchart of the mass spectrometer resolution improvement method based on data analysis according to the present invention.

[0052] Figure 2 This is a schematic diagram of the module of the mass spectrometer resolution improvement system based on data analysis of the present invention. Detailed Implementation

[0053] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0054] This invention provides, for example Figure 1 The data analysis-based mass spectrometer resolution improvement method shown includes the following steps:

[0055] S100 tracks high-energy phase changes point by point in continuously sampled signals with fine time granularity, and connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation.

[0056] To achieve high-precision continuous tracking of phase changes in the high-energy signal segment of a mass spectrometer, the phase state needs to be analyzed point-by-point with fine time granularity in the original sampled signal. This ensures that the rising and falling segments of the signal are completely concatenated into a resolvable continuous phase path, forming a phase fluctuation time band with traceable capabilities. The specific steps are as follows:

[0057] During the signal sampling phase, the time sampling interval is set to the sub-microsecond level, such as 0.5 microseconds, to capture every minute phase change during ion flight. At this sampling rate, at least 200 discrete sampling points can be obtained for a high-energy band lasting 100 microseconds within a mass spectrometry cycle. For each of these sampling points, the amplitude difference of the signal relative to the previous moment is calculated to determine whether it is in the rising phase, falling phase, or plateau region in the time series. Whenever multiple consecutive points show an upward trend in amplitude, it can be identified as a phase rising segment; otherwise, it is a phase falling segment. During this process, no filtering or sampling compression is performed on the signal to ensure that the original form remains intact. The absolute timestamp of the sampling point in the entire sampling period should also be recorded during the sampling process, such as the 35th microsecond, the 36th microsecond, etc., so as to facilitate subsequent chaining in chronological order.

[0058] By sequentially connecting the rising and falling phase segments in time, a continuous phase evolution chain is constructed. During the concatenation process, single-point or double-point amplitude stability regions are allowed between each pair of rising segments and adjacent falling segments, serving as natural transition regions for phase fluctuations. For example, if a rising segment with gradually increasing amplitude appears between 40 and 47 microseconds, and a falling segment with decreasing amplitude immediately appears between 48 and 55 microseconds, with only a brief plateau period between 47 and 48 microseconds in between, the phase change can be considered to have natural continuity. After all rising and falling segments that satisfy temporal continuity and trend connection are sequentially spliced ​​together, a complete phase fluctuation path is formed, which is reflected on the time axis as multiple alternating rising and falling regions, reflecting the transition trajectory of ion energy in the high-energy range.

[0059] After the phase fluctuation path is constructed, each continuous phase path segment is labeled with intervals based on its duration and amplitude range. The duration characterizes whether the phase fluctuation is sufficiently stable, while the amplitude range is used to check for extreme abrupt increases. For example, if the duration of a fluctuation interval is less than 2 microseconds or the amplitude change is less than 0.1 voltage units, it is considered an unstable path segment and will be removed from subsequent analysis intervals; conversely, if the duration exceeds 10 microseconds and the amplitude jump exceeds 0.8 voltage units, the segment will be given high weight and used for focused analysis. Stable phase fluctuation segments selected in this way will be further encoded into traceable paths with complete start and end markers in time, forming traceable phase regions with boundary constraints.

[0060] Within all the selected stable phase fluctuation intervals, a set of phase fluctuation time bands is constructed using the start and end timestamps as references. This time band set begins with the start time and ends with the end time, recording the original amplitude, derivative values, and local extrema of all data points within the period. Taking the period from 60 microseconds to 85 microseconds as an example, this time segment contains over 20 valid sampling points, with the 65th microsecond being the peak of the maximum value and the 75th microsecond being the secondary descent critical point. Embedding this structural information as a multidimensional feature of the time band into the time axis helps in subsequent tracking of fracture locations, determining abrupt boundary changes, and supporting peak structure reconstruction. Each constructed phase fluctuation time band includes time interval information, trend direction, energy change amplitude, and internal key point identifiers, possessing quantifiable and traceable signal tracking capabilities, providing a complete temporal foundation for the next stage of locating fracture structures and constructing crack fingerprint maps.

[0061] S200 compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a list of jump candidate segments in chronological order. The start point, end point and energy surge position are recorded in each jump candidate segment list for subsequent crack location.

[0062] To accurately identify and structurally extract potential chain break locations within the phase fluctuation time band, it is necessary to perform segment-by-segment sampling beat difference analysis on a fine-grained time axis based on the phase fluctuation time band constructed in the previous stage. This allows for the location of abrupt jump anomalies and the aggregation of candidate segments, providing boundary information for subsequent spatial localization and compensation of crack structures. The specific steps are as follows:

[0063] Within each phase fluctuation time band, the phase change amplitude between adjacent sampling beats is calculated point by point according to the sampling time sequence, and the amplitude difference is recorded. Taking the starting point of a phase fluctuation time band as an example, assuming the starting point is the 100th microsecond, the corresponding amplitude is 2.4 voltage units, and the next sampling point is the 100.5th microsecond, the corresponding amplitude is 2.5 voltage units, then the beat difference is 0.1 voltage units. The same operation is performed on subsequent data points such as the 101st microsecond, 101.5th microsecond, etc., to obtain a sequence of adjacent beat differences. To eliminate the influence of local micro-amplitude jitter, a dynamic difference threshold range needs to be set. For example, this threshold range can be dynamically set based on three times the standard deviation of the average beat difference in the phase fluctuation time band. If a beat difference suddenly exceeds this threshold range, it is marked as a suspicious jump point. In a typical data segment, if the difference between two consecutive sampling points reaches 0.9 voltage units, while the difference between adjacent normal points does not exceed 0.2 voltage units, then a discontinuous phase jump can be considered to have occurred, providing an initial set of abnormal points for subsequent aggregation.

[0064] For suspicious jump points, micro-region analysis is performed chronologically within the intervals before and after them to identify the upper and lower boundaries of the anomalous jumps. Points exhibiting both temporal continuity and significant jump amplitude are included in the jump segment. Taking 105 microseconds as the jump center point as an example, if there is a sudden increase in the difference between 104.5 and 105.5 microseconds, and the total span of this segment does not exceed 2 microseconds, then this segment can be used as the start and end time of the initial jump interval. Simultaneously, the points of maximum energy increase and decrease in this data segment need to be retrieved. For example, if the amplitude jumps to 3.2 voltage units at 105.1 microseconds and decreases to 2.6 voltage units at 105.4 microseconds, these are respectively used as the energy surge and fall characteristic points of this jump segment, characterizing the amplitude range and central trend of the jump fluctuation. This time-amplitude dual cross-validation method ensures that the jump segment possesses real physical meaning and continuous anomalous characteristics.

[0065] After identifying multiple jump segments, these segments need to be numbered sequentially according to their temporal position within the phase fluctuation time band and included in a unified list of jump candidate segments. Each list entry should contain four key pieces of information: start time, end time, jump duration, and the location of the point with the maximum jump amplitude. For example, a list entry could have the following structure: start at 118 microseconds, end at 121 microseconds, jump duration of 3 microseconds, maximum jump point at 119.2 microseconds, and peak amplitude of 3.6 voltage units. This structured organization not only improves data organization efficiency but also facilitates subsequent cross-axis analysis and spectrum reconstruction. During this process, all jump segments must originate from the stable phase fluctuation time bands confirmed in the previous stage to ensure temporal continuity and comparability. If a jump segment spans multiple fluctuation segments, it needs to be segmented according to the main sequence path and included in the list separately to prevent abnormal data diffusion.

[0066] The list of candidate jump segments is linked one-to-one with the complete phase fluctuation time band constructed in the previous stage to confirm the location and adjacency of each jump segment in the original signal. After the assignment is completed, each jump segment is assigned a phase evolution path number to form continuous path nodes during the biaxial mapping process. For example, if a jump segment belongs to the middle region of the third phase fluctuation path, it will be marked as 3-P2 in the subsequent crack localization, indicating the second jump point of the third path. This marking method can indicate both its vertical path relationship and its specific location in the original time band. In addition, the original amplitude sequence within the time span of each segment needs to be added to the list to support the reference fitting operation in the subsequent time-series buffering processing. Through the above process, all jump segments are organized in a highly structured and complete traceability manner, laying the data foundation for the next step of constructing a time- and mass biaxial crack map.

[0067] S300 maps each jump fragment to both the time axis and the mass axis based on the list of jump candidate fragments, generating a phase crack fingerprint in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference.

[0068] To achieve accurate multidimensional structural analysis of abrupt jump anomalies, it is necessary to map and fuse each abrupt jump segment in both time and mass physical dimensions based on the obtained list of candidate segments. This will construct a phase crack fingerprint map covering the entire abrupt jump segment, providing a spatial reference for subsequent accurate compensation and sampling adjustments. The specific steps are as follows:

[0069] Based on the start and end times of each segment recorded in the candidate segment list, the corresponding phase abrupt change interval is located on the time axis, and the original signal intensity and amplitude trend changes of all sampling points within that time interval are extracted simultaneously. For example, if the start of a certain jump segment is at 205 microseconds and the end is at 209 microseconds, then all sampling points within these 5 microsecond intervals are extracted from the complete signal sequence, for example, one point every 0.5 microseconds, for a total of 10 sample points. Simultaneously, the amplitude extremes, average rate of rise, peak positions, and differences from adjacent beats of these 10 points are calculated to characterize the transient characteristics of signal changes within that time segment. These indicators form a description of the anomalous characteristics of the jump segment in the time dimension, clearly revealing key information such as its duration, the steepness of signal changes, and waveform breakage trends.

[0070] After reconstructing the signal on the time axis, the jump fragments need to be mapped to the mass axis interval according to their time markers, and the changes in mass number recorded by the mass spectrometer within this time period are extracted. Since there is a deterministic physical mapping relationship between mass number and ion flight time, the corresponding mass number can be deduced from the flight time corresponding to each sampling point between 205 microseconds and 209 microseconds. Assuming that ion flight time is proportional to the square root of the mass number, if 205 microseconds corresponds to a mass number of 120, then 209 microseconds may correspond to a mass number of 122.5. Therefore, the distribution range of the jump fragment on the mass axis can be determined to be mass numbers from 120 to 122.5. Within this mass interval, structural information such as the number of peaks, peak spacing, peak height ratio, and signal-to-noise ratio on the mass spectrum within this range is further extracted to construct an anomaly fingerprint for the mass dimension. This dimension is used to confirm whether there are mass structural anomalies within the jump time period, such as peak overlap, peak breakage, or drift.

[0071] By horizontally fusing anomalous information from the time and mass dimensions, each jump segment is depicted as a crack structure strip with range, direction, and intensity in a two-dimensional coordinate space, with time as the horizontal axis and mass number as the vertical axis. Taking jump segment T3 as an example, if it extends from 210 to 214 microseconds on the time axis and from 125 to 127 mass numbers on the mass axis, a rectangular strip with a width of 4 microseconds and a height of 2 mass numbers can be constructed in the two-dimensional plane. The color intensity of the strip represents the jump intensity, and the thickness of the boundary indicates the degree of signal change. If a peak phase shift greater than 0.6 voltage units or a peak height abrupt change exceeding 50% occurs within this strip, the crack region can be marked as a high-risk anomalous structure. Multiple jump strips combine to form a crack set region, where cracks may exhibit series, intersection, or branching structures. The two-dimensional crack set constructed in this way not only shows the specific location of the jump in time and mass space but also provides a clear target for spatial intervention and local compensation.

[0072] To ensure the consistency and localizability of the crack fingerprint across the entire spectrum, all abrupt bands are numbered chronologically and uniformly embedded into the original spectral spatial coordinate system, forming a cross-axis continuous reference structure. Each crack band simultaneously records its starting coordinates, coverage area, fingerprint intensity score, and relative position within the entire spectrum. For example, crack F5 is located in spectral band 15, between 220 and 224 microseconds, corresponding to a mass number range of 128 to 130, and is rated as a high-intensity crack. For the spatial continuity between adjacent cracks, their overlap rate on the time and mass axes is calculated; if it exceeds 30%, it is marked as a potential crack cluster. The crack fingerprint constructed in this way possesses four attributes: time scale, mass density, intensity distribution, and spatial location. This provides accurate boundary criteria for subsequently inserting sampling adjustment intervals on both sides of the crack and can be used to guide the spatial point selection process for peak shape recovery and phase conditioning, achieving simultaneous optimization of sampling density and spectral quality.

[0073] S400 sets fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map expansion range. Within the adjustment band, the sampling interval and detector gain are fine-tuned so that the phase at both ends of the crack gradually enters the splicable range, forming a time buffer band covering the crack boundary.

[0074] To effectively repair the phase crack boundary, it is necessary to introduce time-continuous and quality-corresponding adjustment regions at both ends of the constructed phase crack fingerprint. By finely controlling the sampling interval and detection gain, the signal on both sides is gradually returned to a stitchable state, ultimately constructing a temporal buffer band covering the crack boundary. The specific steps are as follows:

[0075] Based on the start and end points of the time axis for each crack strip in the generated phase crack fingerprint, its extension range to both sides is calculated, and extension boundaries for adjustment are set. For example, if a crack strip covers a time interval from 312 microseconds to 316 microseconds and a quality range from 135 to 137 mass numbers, then an adjustment boundary can be extended by 2 microseconds before and after it, forming an adjustment interval starting at 310 microseconds and ending at 318 microseconds. Within this interval, time is divided every 0.25 microseconds, and each division point is marked as a potential sampling adjustment point. Edge points overlapping with crack strips are used as high-priority fine-tuning points, and points farther away have progressively lower priority. This method provides clear transition zone boundaries at both ends of the crack, providing an operable range for subsequent time-granular sampling control operations.

[0076] Within the adjustment range, a fine-tuning strategy for the sampling interval is set based on the distance of each sampling point from the crack center, adjusting the sampling period point by point to buffer abrupt phase changes. With the crack center as the axis of symmetry, the closer a sampling point is to the center, the smaller its sampling period adjustment range, typically between 0.01 and 0.03 microseconds; points farther from the center can be set to an adjustment range of 0.05 microseconds. For example, at the 311th microsecond on the left side of the crack, the original sampling interval is 0.5 microseconds, adjusted to 0.48 microseconds; at the 310.5th microsecond, it is adjusted to 0.47 microseconds; and at the 317.5th microsecond on the right side of the crack, it is adjusted from 0.5 microseconds to 0.53 microseconds. Through this progressive fine-tuning mechanism, the phase data points on both sides of the crack are gradually compressed or expanded along the time axis, allowing the originally discontinuous phase trend to obtain a transition gradient in the sampling dimension. This adjustment path prevents abrupt jumps in the signal at the crack edge, instead forming a buffered phase transition trend.

[0077] While fine-tuning the sampling period, the amplitude detection sensitivity of each sampling point within the adjustment interval is also fine-tuned. This involves controlling the detector gain to enhance the effective identification capability of the crack edge signal. Specifically, if the original amplitude of a sampling point is 1.8 voltage units, but the amplitude of an adjacent point suddenly jumps to 3.2 voltage units, there is a significant risk of abrupt change. In this case, the detector gain is adjusted from the standard value of 1.0 to 1.2 or higher, making the response to small phase changes more sensitive and thus improving the signal-to-noise ratio of the sampling point in the crack transition zone. Taking the 311th microsecond on the left side of the crack as an example, the detector gain is adjusted from the default value of 1.0 to 1.15; at 310.5 microseconds, it is adjusted to 1.2; and at 317 microseconds on the right side of the crack, it is adjusted to 0.95, narrowing its signal response range and suppressing the influence of excessive signal fluctuations. This gradual gain adjustment method on both sides ensures that the signal reaches the expected amplitude change trend before entering the crack center region, thereby establishing a stable detection transition zone.

[0078] The time intervals, after dual adjustments to the sampling period and detection gain, are integrated into a complete buffer band structure, serving as a bridge for the transition from the crack region to the continuous phase path. This buffer band not only maintains continuity on the time axis but also preserves the symmetry and fitability of the peak structure on the mass axis. For example, the buffer band from 310 microseconds to 318 microseconds contains 32 high-density sampling points, each with its own sampling time, amplitude response value, and adjusted gain parameters, forming the smallest continuous unit sequence usable for peak reconstruction. The buffer band is connected to the original signal at both ends, and its central region overlaps with the crack strip, forming a stable signal transition band. In subsequent phase conditioning, the buffer band serves as the supporting basis for interpolation and rearrangement, ensuring that the phase is progressively aligned and smoothly transitioned at the crack boundary, guaranteeing the continuity and computability of the final peak shape in both spatial and temporal dimensions.

[0079] S500 performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

[0080] To achieve complete peak shape recovery after passing through the crack region and ensure its integrability under continuous scanning, it is necessary to rely on the time-series buffer band constructed in the previous stage and perform directional and hierarchical rearrangement operations within it. Through a spiral path-based micro-offset strategy, the phase is gradually aligned and smoothly reshaped, allowing the signal to return to a stable state that is fitable, integrable, and quantifiable. The specific steps are as follows:

[0081] Based on the crack length information recorded within the buffer zone, the starting center point and the number of path expansion layers for the spiral rearrangement are determined. The crack lengths within the buffer zone provide the basic geometric scale for executing the rearrangement path. Taking a crack segment as an example, if its span is from the 405th microsecond to the 410th microsecond and extends along the mass axis from mass number 140 to 143, then the actual span of the crack can be converted into a time width of 5 microseconds and a height of 3 mass units. Based on this span, the rearrangement center is defined within the buffer zone with the crack center point (407.5 microseconds, mass number 141.5), and the outer expansion boundary is defined as ±1.5 microseconds and ±1.5 mass units. This area is then divided into several equidistant grid points. These grid points will serve as the node sequence for the spiral path, and according to the radiation direction of the polar coordinate system, the path arrangement sequence is generated sequentially from the inside out, thus forming a concentric spiral rearrangement skeleton around the crack center.

[0082] Along the spiral path, a corresponding micro-offset command sequence is assigned to each grid node. This command sequence contains offset values ​​in both the time and mass axes, exhibiting a hierarchical increasing characteristic. Taking the first loop of the spiral path as an example, it contains eight nodes, with the time offset between nodes controlled within ±0.02 microseconds and the mass offset within ±0.03 mass units. The second loop expands to twelve nodes, with the time offset adjusted to ±0.04 microseconds and the mass offset to ±0.06 mass units. This progressive offset structure allows nodes closer to the crack boundary to have a higher adjustment range, enabling buffering and reconstruction of the strong abrupt changes at the crack center, while peripheral nodes complete fine corrections to the signal trajectory with smaller offsets. The micro-offset values ​​of all nodes are determined by the original amplitude trend in the buffer zone and the degree of change in adjacent beats, giving the entire rearrangement path dynamic adaptability.

[0083] After determining the offset command for each node, a micro-displacement operation is performed point-by-point to rearrange the sampling points within the crack region to the new coordinate positions defined by the spiral path. During execution, the original time and quality positions of the node's sampling points are used as initial references to move the corresponding data points to the offset positions, and the new sampling timestamps and quality number labels are recorded. For example, a data point originally located at 406 microseconds with a quality number of 141 is repositioned to 406.03 microseconds with a quality number of 140.97 after offset; a data point originally located at 409 microseconds with a quality number of 142.5 is repositioned to 408.94 microseconds with a quality number of 142.6. All offsets are performed within the original data boundaries to ensure that the overall signal framework is not destroyed, while maintaining the natural connectivity between data points during the rearrangement process. After all data points in the entire buffer band have been rearranged, the waveform structure of the crack region is broken down and reorganized, so that the original phase jumps and breaks are no longer concentrated, but distributed to various regions of the spiral path, exhibiting a continuously changing and smoothly fitting trend.

[0084] After rearrangement, the entire peak shape is smoothed. Using the repositioned data points along the spiral path as a base, secondary fine-tuning is performed to enhance the fit and waveform consistency between adjacent data points. This smoothing process does not introduce external interpolation data; it only makes minor corrections to the position and amplitude of existing offset points, transforming potential data jumps caused by rearrangement into continuous transitions. For example, if a trough is excessively steep between 407 and 408 microseconds at key points on either side of the crack boundary, the amplitude difference between the two points is gradually reduced from 1.2 voltage units to 0.6 voltage units, smoothing the curve and eliminating abrupt changes. Throughout the buffer zone, the rearrangement offset and smoothing correction, decreasing outwards from the crack center, work together to gradually restore the peak shape to a complete and fitable symmetrical form on both the time and mass axes, ultimately forming a stable peak array with a complete structure that can be integrated. This provides a reliable foundation for quantitative detection, mass calibration, and signal identification in subsequent mass spectrometry data analysis.

[0085] This invention establishes a continuous phase tracking foundation by constructing a phase fluctuation time band, then accurately extracts abrupt jumps and maps them to a two-dimensional space of time and mass axes, forming a crack fingerprint map with directional and spatial coverage characteristics. Combined with fine-grained sampling control and a spiral rearrangement phase conditioning process, dynamic repair and peak reshaping of anomalous signal regions are achieved. Ultimately, in continuous scanning mode, the integrability of peaks and the structural integrity of the spectrum are restored, significantly improving the resolution, stability, and quantitative accuracy of mass spectrometry analysis, expanding the performance ceiling of traditional mass spectrometers, and providing a highly reliable hardware and software integrated enhancement method for high-precision mass spectrometry detection tasks.

[0086] This invention provides, for example Figure 2 The mass spectrometer resolution enhancement system based on data analysis shown includes a phase fluctuation construction module, a jump fragment extraction module, a crack fingerprint generation module, a temporal buffer band construction module, and a helical rearrangement training module.

[0087] The phase fluctuation construction module tracks the phase changes of the high-energy band point by point in the continuously sampled signal with fine time granularity. It connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation.

[0088] The jump segment extraction module compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a jump candidate segment list in chronological order. The starting point, ending point and energy surge position are recorded in each jump candidate segment list for subsequent crack location.

[0089] The crack fingerprint generation module maps each jump fragment to the time axis and mass axis simultaneously based on the list of jump candidate fragments, generating a phase crack fingerprint map in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference.

[0090] The timing buffer band construction module sets up fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map extension range. Within the adjustment band, the sampling interval fine-tuning and the detector gain fine-tuning are performed so that the phase at both ends of the crack gradually enters the splicable interval, forming a timing buffer band covering the crack boundary.

[0091] The spiral rearrangement training module performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped step by step in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

[0092] The mass spectrometer resolution enhancement method based on data analysis provided in this embodiment of the invention is implemented through the aforementioned mass spectrometer resolution enhancement system based on data analysis. For details of the specific methods and processes of the mass spectrometer resolution enhancement system based on data analysis, please refer to the embodiments of the mass spectrometer resolution enhancement method based on data analysis described above, which will not be repeated here.

[0093] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for improving mass spectrometer resolution based on data analysis, characterized in that, Includes the following steps: S100 tracks high-energy phase changes point by point in continuously sampled signals with fine time granularity, and connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation. S200 compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a list of jump candidate segments in chronological order. The start point, end point and energy surge position are recorded in each jump candidate segment list for subsequent crack location. S300 maps each jump fragment to both the time axis and the mass axis based on the list of jump candidate fragments, generating a phase crack fingerprint in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference. S400 sets fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map expansion range. Within the adjustment band, the sampling interval and detector gain are fine-tuned so that the phase at both ends of the crack gradually enters the splicable range, forming a time buffer band covering the crack boundary. The S500 performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped step by step in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

2. The method for improving mass spectrometer resolution based on data analysis according to claim 1, characterized in that, Step S100 includes: During the signal sampling stage, the continuous sampling signal is sampled with fine time granularity. The amplitude difference value of each point is calculated and the absolute timestamp is recorded to form the phase rising segment and the phase falling segment. After the phase rising segment and the phase falling segment are formed, the phase rising segment and the phase falling segment are connected in chronological order and an amplitude stabilization region is set to form a continuous phase evolution chain. After forming a continuous phase evolution chain, the continuous phase path is labeled with intervals according to the duration and amplitude range to filter out stable phase fluctuation intervals. After selecting stable phase fluctuation intervals, the original amplitude, derivative value and local extreme state are recorded with reference to the time marker of the stable phase fluctuation interval to form a set of phase fluctuation time bands.

3. The method for improving mass spectrometer resolution based on data analysis according to claim 2, characterized in that, Step S200 includes: In the phase fluctuation time band, the phase change amplitude between adjacent sampling beats is calculated point by point according to the sampling time sequence, and the amplitude difference is recorded to form a sequence of adjacent beat difference values. After forming the difference sequence of adjacent beats, micro-region analysis is performed on the amplitude abrupt increase points in the difference sequence of adjacent beats in chronological order to identify abrupt jump segments with temporal continuity and significant jump amplitude; After identifying the jump segments, the jump segments are numbered in chronological order and the start time, end time, jump duration, and the location of the point with the maximum jump amplitude are recorded to form a list of candidate jump segments. After forming a list of candidate jump segments, the list of candidate jump segments is associated with the phase fluctuation time zone, and the original amplitude sequence of the jump segments within the time span is recorded to form a data foundation with complete traceability capability.

4. The method for improving mass spectrometer resolution based on data analysis according to claim 3, characterized in that, After forming a list of candidate jump segments, the amplitude change trend of the sampling points before and after the jump segment is checked by using the start time and end time recorded in the list of candidate jump segments as the boundary. After the continuity check is completed, the amplitude gradient change inside the jump segment is analyzed with the position of the point with the maximum jump amplitude as the center, so as to confirm the abnormal intensity range of the jump segment in the phase fluctuation time band and use it for key weight marking in subsequent biaxial mapping.

5. The method for improving mass spectrometer resolution based on data analysis according to claim 3, characterized in that, Step S300 includes: Based on the start and end times of each segment recorded in the list of candidate segments for sudden jumps, the phase jump interval is located on the time axis, and the original signal intensity and amplitude trend changes within the phase jump interval are extracted to form time dimension anomalous features. After forming time-dimensional anomalous features, the jump segments are mapped to the mass axis interval according to the time marker, and the number of spectral peaks, peak spacing, peak height ratio and signal-to-noise ratio within the mass interval are extracted to form quality-dimensional anomalous features. After forming the quality dimension anomaly features, the time dimension anomaly features are fused with the quality dimension anomaly features to depict crack structure stripes with range, direction and intensity in two-dimensional coordinate space; After the crack structure strips are formed, they are embedded into the spectral spatial coordinate system in chronological order, and the starting coordinates, coverage area and fingerprint intensity score are recorded to form a phase crack fingerprint map.

6. The method for improving mass spectrometer resolution based on data analysis according to claim 5, characterized in that, After the phase crack fingerprint map is formed, the temporal overlap rate and mass overlap rate of adjacent crack structure strips in the phase crack fingerprint map are compared. When the overlap rate exceeds the preset overlap threshold, the adjacent crack structure strips are merged into crack clusters to form a continuous abnormal region. After the continuous abnormal region is formed, the time start point, time end point, mass span and fingerprint intensity of the continuous abnormal region are recorded to improve the crack location accuracy.

7. The method for improving mass spectrometer resolution based on data analysis according to claim 5, characterized in that, Step S400 includes: Based on the start and end points of the time axis of the crack strips in the phase crack fingerprint image, an extension range is set, and potential sampling adjustment points are divided according to the extension range to form an adjustment interval; After the adjustment interval is formed, the sampling interval is finely adjusted point by point according to the distance between each sampling point and the crack center so that the phase change forms a transition gradient on the time axis; After forming the transition gradient, the detector gain is fine-tuned at each sampling point within the adjustment interval to enhance the effective identification capability of crack edge signals and form a detection transition zone. After forming the detection transition band, the time intervals processed by sampling interval fine-tuning and detector gain fine-tuning are integrated into a continuous buffer band structure to form a stable transition between the crack region and the continuous phase path.

8. The method for improving mass spectrometer resolution based on data analysis according to claim 7, characterized in that, Within the adjustment range, the sampling time and amplitude response of the sampling points are synchronously calibrated with the crack center as the reference so that the sampling points inside the buffer zone are distributed in a progressive and continuous manner. After synchronous calibration, the sampling points inside the buffer zone are locally smoothed according to the changing trend of the amplitude response so that the phases at both ends of the buffer zone are gradually aligned and a continuous peak transition path is formed.

9. The mass spectrometer resolution improvement method based on data analysis according to claim 7, characterized in that, Step S500 includes: The starting center point and path expansion layers of the spiral rearrangement are determined based on the crack length recorded inside the temporal buffer band, and the outer expansion boundary is delineated to form the spiral rearrangement skeleton. After forming the spiral rearrangement skeleton, each grid node is assigned a micro offset instruction sequence with hierarchical increments along the spiral path so that the node has time direction offset and mass direction offset; After forming the micro-offset instruction sequence, perform micro-displacement operations point by point to rearrange the sampling points in the crack area into the spiral path and record the new sampling timestamp and quality number label; After forming the spiral path, full-segment smoothing is performed on the repositioned data points along the spiral path to gradually restore the continuous structure of the peak shape on the time axis and the quality axis.

10. A mass spectrometer resolution enhancement system based on data analysis, used to implement the mass spectrometer resolution enhancement method based on data analysis as described in any one of claims 1-9, characterized in that, It includes a phase fluctuation construction module, a jump fragment extraction module, a crack fingerprint generation module, a temporal buffer band construction module, and a spiral rearrangement training module: The phase fluctuation construction module tracks the phase changes of the high-energy band point by point in the continuously sampled signal with fine time granularity. It connects each phase rise and phase fall in time sequence to generate a continuously extended phase fluctuation time band, which is used to build a comprehensive phase tracking foundation. The jump segment extraction module compares the difference between adjacent sampling beats segment by segment around the phase fluctuation time band, and gathers the positions where phase chain breaks occur into a jump candidate segment list in chronological order. The starting point, ending point and energy surge position are recorded in each jump candidate segment list for subsequent crack location. The crack fingerprint generation module maps each jump fragment to the time axis and mass axis simultaneously based on the list of jump candidate fragments, generating a phase crack fingerprint map in the dual-axis space that characterizes the crack extension direction and diffusion morphology, which is used to form a cross-axis continuous anomalous feature reference. The timing buffer band construction module sets up fine-grained sampling beat adjustment bands on both sides of the phase crack fingerprint map according to the fingerprint map extension range. Within the adjustment band, the sampling interval fine-tuning and the detector gain fine-tuning are performed so that the phase at both ends of the crack gradually enters the splicable interval, forming a timing buffer band covering the crack boundary. The spiral rearrangement training module performs a spiral rearrangement phase training process within the time buffer zone based on the time buffer zone. It generates a multi-level micro-offset instruction series according to the crack length recorded by the time buffer zone, so that the peak shape is aligned and smoothly reshaped step by step in the spiral rearrangement path, thereby restoring the stable integrable peak series in continuous scanning.

Images