Integrated analysis method and apparatus for multimodal retinal ganglion cell data

Abstract

Provided are an integrated analysis method and apparatus for marking multimodal retinal ganglion cell (RGC) data of central nervous system cell aging. The method uses an aged RGC as a biomarker of the central nervous system cell aging, and comprises: applying an ETDRS grid to a reconstructed plan view of a macular region in an OCT retinal thickness image, so as to divide the plane view into a plurality of regions, wherein the OCT thickness image comprises thickness and volume data values of five different retinal layers; grouping the plurality of regions and determining the average thickness and volume data distribution of each region to measure the abundance of RGCs; using a posterior pole module to record an OCTA image and performing fine stratification to obtain a retinal microvascular change in the OCTA image; and obtaining 103 hexagonal mfERG reports where pERG and ETDRS grids have a same center at a macula. On the basis of the obtained information, the histological and functional analysis of the RGCs is combined with the precise position distribution and corresponding blood supply microvascular changes to provide a non-invasive early detection analysis module that reflects the health of the RGCs.

Description

Method and device for integrated analysis of multimodal retinal ganglion cell data

[0001] This application claims priority to U.S. Provisional Application No. 63 / 579,170, filed on August 28, 2023, the disclosure of which is incorporated herein by reference in its entirety as a part of this application. Background Art

[0002] In vivo tracing of mammalian central nervous system (CNS) neurons can be achieved through anterograde or retrograde delivery of specific dyes via nerve fibers or viral delivery of genetically manipulated fluorescent tags, followed by observation through micropores using specialized fluorescence microscope lenses. Noninvasive and sustainable tracking of specific CNS neuron types to monitor CNS aging has yet to be reported. Studies have shown that patients with mood disorders exhibit underlying structural changes in the brain. For example, Quiroz and Manji (2002) found a significant decrease in the volume of certain CNS regions, as well as in the number and / or size of glial cells and neurons. However, direct measurement of CNS volume or the abundance of glial cells and neurons in living humans remains difficult. The retina, as an extracranial extension of brain neurons and the only directly observable neurovascular tissue, provides an ideal window for monitoring CNS neuronal properties. Because aging is a natural process of neuronal death and decline in the human brain, studying the changes in the properties of retinal ganglion cells (RGCs) during CNS cell aging is intended to identify retinal biomarkers reflecting CNS neuronal aging and provide noninvasive analytical methods for early detection.

[0003] Optical coherence tomography (OCT) is a non-invasive imaging technique that uses light waves to perform cross-sectional scans of the retina. It is based on the concept of interferometry to create cross-sectional images of the retina with an accuracy of at least 10-15 microns (Costa et al., 2006). OCT-based angiography (OCTA) is a new non-invasive method for visualizing the retinal vasculature. It is based on the principle of diffraction particle movement of moving red blood cells to determine the location of blood vessels passing through various ocular segments without the need for any intravascular dye (Le & Patel, 2020), analyzing scattering from static background tissue to create high-resolution angiography. OCTA technology allows imaging of blood flowing in the retina and displays the choroidal vasculature through coronal and depth-encoding slabs, presented together with structural OCT B-scans. OCTA can be performed on the same equipment as OCT. Together, they provide detailed flow imaging of the deep retinal vascular plexus and choriocapillaris, as well as retinal structures in different layers. To measure the electrical activity of the retina in response to light stimulation, diagnostic tests such as electroretinography (ERG) can be used. There are three main types of ERG tests: pattern ERG (pERG), flash ERG (fERG), and multifocal ERG (mfERG). The pERG is commonly used as a test for conditions affecting ganglion cells and various neurodegenerative conditions; the mfERG shows the condition of the retina as a whole; and the mfERG is best used to measure dysfunction in the central retinal area. Using different types of ERG tests allows for a comprehensive assessment of a patient's retinal function, reliable monitoring of changes in retinal function over time, and evaluation of treatment effectiveness.

[0004] However, these studies have primarily focused on changes in the retina itself, lacking correlation analysis with specific neurons in the CNS. Currently, no standardized model for comprehensive analysis of OCT, OCTA, and ERG results has been found, nor has an effective biomarker for objectively and quantitatively assessing the status of CNS neurons been found.

[0005] Summary of the Invention

[0006] Embodiments of the present invention relate to methods for the integrated analysis of retinal structural and functional biomarkers for identifying CNS cell senescence.

[0007] According to one embodiment of the present invention, a method for integrating and analyzing multimodal retinal ganglion cell data for central nervous system marker cell aging is provided, wherein RGC aging is used as a biomarker for central nervous system cell aging. The method comprises: applying an Early Treatment of Diabetic Retinopathy Study (ETDRS) grid to a macular plan view reconstructed from an optical coherence tomography (OCT) thickness map, dividing the macular plan view into a plurality of regions, wherein the OCT thickness map comprises data values ​​of thickness and volume of five different retinal layers, wherein the five different retinal layers comprise nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), nerve fiber layer to inner plexiform layer (NFL-IPL) and ganglion cell layer to inner plexiform layer (GCL-IPL); grouping the plurality of regions, and determining the average thickness and volume data distribution of each region of the five different layers to detect the abundance of RGCs; recording optical tomographic angiography (OCTA) using a posterior pole module and finely stratifying to obtain retinal microvascular changes of the OCTA; obtaining pER G reports and analyzes the latency of N35, P50, N95, and the amplitudes of P50 (N35-P50) and N95 (P50-N95); obtains a multifocal electroretinogram (mfERG) report of 103 hexagons with the same center of the ETDRS grid at the macula, and the mfERG report includes analysis of the latency of P1 and N1, and the amplitudes of P1 and N1; and identifies whether retinal ganglion cells (RGCs) are aging based on the average thickness and volume data of each area of ​​the five different layers of the OCT thickness map, retinal microvascular changes of the OCTA, the latency of N35, P50, N95, and the amplitudes of P50 (N35-P50) and N95 (P50-N95) in the pERG report, and the latency of P1 and N1, and the amplitudes of P1 and N1 in the mfERG report, thereby identifying cell aging in the central nervous system. Combining RGC histological and functional analysis with precise location and distribution as well as corresponding changes in blood supply microvasculature provides a non-invasive early detection analysis module reflecting RGC health.

[0008] According to an embodiment of the present invention, the method further includes: using the posterior pole module to record the reconstructed macular area plan view in the OCT thickness map.

[0009] According to an embodiment of the present invention, dividing the macular area plan view into a plurality of regions comprises: dividing the macular area plan view into eight regions, the eight regions comprising an inner ring superior, an inner ring nasal, an inner ring inferior, an inner ring temporal, an outer ring superior, an outer ring nasal, an outer ring inferior, and an outer ring temporal; and

[0010] Grouping the plurality of regions comprises: grouping the macular region to form three annular regions comprising: a fovea having a diameter within 1 mm; an inner annular zone having a diameter between 1 mm and 3 mm; and an outer annular zone having a diameter between 3 mm and 6 mm, or,

[0011] Group the macular area to form 4 zones including: superior, nasal, inferior, and temporal zones with an annular area between 1 mm and 6 mm, or

[0012] The macular areas were grouped to form 8 regions including: inner annular superior, inner annular nasal, inner annular inferior, and inner annular temporal regions with diameters between 1 mm and 3 mm and outer annular superior, outer annular nasal, outer annular inferior, and outer annular temporal regions with diameters between 3 mm and 6 mm.

[0013] According to an embodiment of the present invention, for grouping the macular area to form 3 areas, determining the thickness and volume data of each area of ​​the five different layers includes: calculating the average thickness and volume of the inner annular zone areas of the five different layers; calculating the average thickness and volume of the outer annular zone areas of the five different layers; and calculating the average thickness and volume of the total annular zone, which is the combined area of ​​the inner annular zone areas and the outer annular zone areas.

[0014] According to an embodiment of the present invention, for grouping the macular area to form 4 areas, determining the thickness and volume data of each area of ​​the five different layers includes: calculating the average thickness and volume of the total upper area, wherein the total upper area is the combination area of ​​the inner ring upper area and the outer ring upper area; calculating the average thickness and volume of the total nasal area, wherein the total nasal area is the combination area of ​​the inner ring nasal area and the outer ring nasal area; calculating the average thickness and volume of the total lower area, wherein the total lower area is the combination area of ​​the inner ring lower area and the outer ring lower area; and calculating the average thickness and volume of the total temporal area, wherein the total temporal area is the combination area of ​​the inner ring temporal area and the outer ring temporal area.

[0015] According to an embodiment of the present invention, for grouping the macular area to form 8 areas, determining the thickness and volume data of each area of ​​the five different layers includes: calculating the average thickness and volume of the inner ring upper area, inner ring nasal area, inner ring lower area, and inner ring temporal area of ​​the five different layers; and calculating the average thickness and volume of the outer ring upper area, outer ring nasal area, outer ring lower area, and outer ring temporal area of ​​the five different layers.

[0016] According to an embodiment of the present invention, the method also includes extracting an area with a thickness greater than 45 μm from the OCT thickness map and calculating the area of ​​the area by an image processing method; extracting an area within a 6 mm circle from the ETDRS grid and calculating the area of ​​the area by an image processing method; determining a parameter Area based on the ratio of the area of ​​the area greater than 45 μm in the OCT thickness map to the area within the 6 mm circle of the ETDRS grid, wherein the parameter Area is used as a comprehensive indicator of ganglion cell abundance; and determining whether RGC aging exists based on the parameter Area.

[0017] According to an embodiment of the present invention, using the posterior pole module to record optical tomography angiography (OCTA) and finely layering to obtain the retinal microvascular changes of the OCTA also includes: acquiring OCTA vascular image data; extracting the area within a circle with a diameter of 2 mm centered on the macula as the region of interest (ROI) and cropping out the area of ​​the image; calculating the vascular abundance, vascular length, vascular diameter and vascular tortuosity by applying the OCTAVA method to the ROI; and calculating the foveal avascular zone (FAZ) parameters, the FAZ parameters including the FAZ area index and the FAZ roundness index; identifying whether RGC is aging based on the retinal microvascular changes of the OCTA also includes: determining whether central nervous system cell aging is present based on the vascular abundance, vascular length, vascular diameter and vascular tortuosity and the foveal avascular zone (FAZ) parameters.

[0018] According to an embodiment of the present invention, the method also includes performing a 48-minute pattern electroretinogram (pERG) test twice and obtaining a pERG report; and extracting N35, P50, and N95 latency and P50 and N95 amplitude data from the pERG report by an OCR Space text extraction method; and determining whether RGC aging exists based on the N35, P50, and N95 latency and P50 and N95 amplitudes.

[0019] According to an embodiment of the present invention, obtaining an mfERG report includes: performing an mfERG test and obtaining an mfERG report; dividing the 103 hexagons into four different areas / quadrants with a horizontal line and a vertical line, and calculating the average latency and amplitude of P1 and N1 in the four different quadrants, and / or dividing the 103 hexagons into six equally spaced concentric annular areas around the central disk, and calculating the average latency and amplitude of P1 and N1 in the six different rings.

[0020] According to another embodiment of the present invention, a method for a comprehensive analysis module for effective, quantitative and reliable evaluation, identification, assessment and intervention is provided. The method includes analyzing the OCT thickness and volume data of different areas of the nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), nerve fiber layer to inner plexiform layer (NFL-IPL) and ganglion cell layer to inner plexiform layer (GCL-IPL); analyzing the OCT custom parameter Area of ​​the GCL thickness map; analyzing the OCTA superficial and deep vascular plexuses and calculating the vascular abundance, vascular length, vascular diameter, vascular tortuosity and / or foveal avascular zone (FAZ) parameters, including FAZ area index and FAZ roundness index; generating a pERG report and analyzing the N35, P50, N95 latency (implicit time) and amplitudes of P50 (N35-P50) and N95 (P50-N95); obtained mfERG reports and analyzed P1 and N1 latencies and P1 and N1 amplitudes; and performed a comprehensive analysis of retinal structural, vascular, and functional parameters by correlating the Early Treatment of Diabetic Retinopathy Study (ETDRS) grid with 103 hexagons with the same center at the macula in the mfERG reports.

[0021] In another embodiment of the present invention, a method for comprehensive and detailed analysis of OCT thickness and volume is provided by applying three different segmentation methods at the macular area and by valid customized parameters. The method includes: recording the OCT thickness map using a posterior pole module; applying an ETDRS grid to the OCT thickness map to divide the retina into eight regions, the eight regions including inner superior, inner nasal, inner inferior, inner temporal, outer superior, outer nasal, outer inferior, and outer temporal; grouping the eight regions based on three different division methods; calculating the average thickness and volume data of each region in different layers; the different layers include nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), from nerve fiber layer to inner plexiform layer (NFL-IPL), and from ganglion cell layer to inner plexiform layer (GCL-IPL); cropping the OCT thickness map output by the software so that only the right half of the numerical data is retained for OCT thickness and volume analysis; and The Space method extracts numerical data from the cropped OCT thickness map. The first of the three different segmentation methods includes calculating the average thickness and volume of the inner Annulus sector in five different layers; calculating the average thickness and volume of the outer Annulus sector in the five different layers; and calculating the average thickness and volume of the total Annulus, which is the combined area of ​​the inner Annulus sector and the outer Annulus sector in the five different layers. The second of the three different segmentation methods includes calculating the average thickness and volume of the total superior region, which is the combined area of ​​the inner superior and outer superior regions in the five different layers; calculating the average thickness and volume of the total nasal region, which is the combined area of ​​the inner nasal and outer nasal regions in the five different layers; calculating the average thickness and volume of the total inferior region, which is the combined area of ​​the inner inferior and outer inferior regions in the five different layers; and calculating the average thickness and volume of the total temporal region, which is the combined area of ​​the inner temporal region and the outer temporal region in the five different layers. The third of the three different segmentation methods involves calculating the average thickness and volume of the inner annular superior, inner annular nasal, inner annular inferior, and inner annular temporal regions across five different layers; and calculating the average thickness and volume of the outer annular superior, outer annular nasal, outer annular inferior, and outer annular temporal regions across five different layers. The OCT thickness map output by the software was cropped so that only the left half of the numerical data was retained for calculating the custom parameter Area.Regions with a thickness greater than 45 μm were extracted and their areas were calculated using image processing methods. Regions within the 6 mm circle of the ETDRS grid were extracted and their areas were calculated using image processing methods. The value of Area was the ratio of the area with a thickness greater than 45 μm on the OCT thickness map to the area within the 6 mm circle of the ETDRS grid.

[0022] In another embodiment of the present invention, a method for comprehensive and detailed analysis of superficial and deep vascular plexus images of OCTA is provided. The method comprises: recording the OCTA vascular image data; extracting an area within a 2 mm circle centered on the macula and cropping the rest of the image as a trade-off for including more vascular areas without losing vascular details; calculating vascular abundance, vessel length, vessel diameter, and vessel tortuosity by applying the OCTAVA method to the cropped vascular image; and calculating foveal avascular zone (FAZ) parameters, including FAZ area index and FAZ circularity index.

[0023] In certain embodiments of the present invention, a method for quantitatively analyzing a pERG report includes recording a 48-minute pERG test twice and obtaining a pERG report; and extracting numerical data of N35, P50, and N95 latencies and P50 and N95 amplitudes from the pERG report by an OCR Space text extraction method.

[0024] In certain embodiments of the present invention, a method for comprehensively analyzing mfERG reports by dividing the analysis area in two different ways is provided. The method includes recording 103 mfERG tests and obtaining an mfERG report; dividing the 103 hexagonal areas in the test report into four different areas / quadrants with a horizontal line and a vertical line, and calculating the average latency and amplitude of P1 and N1 in the four different quadrants; and dividing the 103 hexagons into six equally spaced concentric eccentricities around a central disk, and calculating the average latency and amplitude of P1 and N1 in six different concentric rings.

[0025] In another embodiment of the present invention, a method for integrated analysis of retinal structure and function biomarkers includes applying an ETDRS grid and 103 hexagons to the macular region, both centered on the fovea; and correlating regional changes in retinal structure and function by correlating the overlapping areas of the two overlays.

[0026] According to another embodiment of the present invention, a computer program product is provided, comprising: a non-transitory computer-executable storage device having computer-readable program instructions embodied thereon, wherein the computer-readable program instructions, when executed by a computer, cause the computer to perform the above method.

[0027] According to another embodiment of the present invention, there is provided an integrated analysis device for multimodal retinal ganglion cell data for marking aging, comprising at least one processor configured to execute the above method. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] FIG1 shows the process of applying three different partitioning methods to different layers of the retina at the macular area to analyze OCT thickness and volume changes.

[0029] FIG2 shows an image of the Early Treatment Diabetic Retinopathy Study (ETDRS) grid, comprising circles having diameters of 1 mm, 3 mm, and 6 mm, respectively, overlaid on an OCT thickness map dividing the OCT thickness map into 9 regions.

[0030] FIG3 shows an exemplary OCT thickness map of the ganglion cell layer (GCL) along with software-generated statistics including the mean thickness and volume of the corresponding eight regions.

[0031] FIG4 illustrates an exemplary use of the OCR Space to extract text from an image, showing the user interface, input, and output.

[0032] Figure 5A shows an analysis method, in which the macular area is divided into three areas, namely, the inner ring zone, the outer ring zone, and the inner ring zone plus the outer ring zone (total ring zone); Figure 5B shows another analysis method, in which the macular area is divided into four areas, namely, the superior side, the nasal side, the inferior side, and the temporal side; Figure 5C shows another analysis method, in which the macular area is divided into nine areas, including the superior inner ring, the nasal inner ring, the inferior inner ring, the temporal inner ring, the superior outer ring, the nasal outer ring, the inferior outer ring, and the temporal outer ring.

[0033] FIG6 shows the original OCT thickness map and the OCT thickness map after image preprocessing in which regions with a thickness exceeding 45 μm are extracted.

[0034] FIG7 shows a schematic diagram of vascular parameters, including vascular area abundance, mean / distribution of vascular diameter, total length / distribution of vascular length, vascular tortuosity, vascular branch point abundance, and vascular fractal abundance.

[0035] FIG8 shows an exemplary superficial vascular plexus image on the left side and an exemplary deep vascular plexus image on the right side.

[0036] FIG9 shows an original blood vessel image and a pre-processed blood vessel image. In the pre-processed blood vessel image, an area within a 2 mm circle centered on the macula is extracted as a further analysis area.

[0037] Figure 10 shows an exemplary use of the OCTAVA toolbox. After a microvascular image is uploaded, the toolbox first performs binarization processing on it to extract the vascular structure from the background, and then applies a skeletonization algorithm to further extract the vascular structure. Finally, the distribution information of the vascular parameters is output in Excel format.

[0038] FIG11 shows an exemplary use of OCR Space to extract useful data from an uploaded mfERG report. The amplitude and latency data of N1 and P1 within 6 rings are extracted for subsequent analysis.

[0039] FIG. 12 shows an exemplary output pERG report from the ERG software.

[0040] Figure 13A shows an analysis method of 103 hexagons at the macular area, wherein the 103 hexagons are divided into 4 quadrants, and the average amplitude and latency of P1 and N1 are analyzed separately for each of the 4 quadrants; Figure 13B shows another analysis method, wherein the 103 hexagons at the macular area are divided into 6 concentric and equidistant eccentricities, and the average amplitude and latency of P1 and N1 are analyzed separately for each of the 6 rings.

[0041] Figure 14 shows a schematic diagram with two overlays: one is a 1, 3, 6 mm ETDRS grid overlay and the other is a 103 hexagonal division with the same center at the macula shown on the same image.

[0042] Figure 15A shows the three distinct layers of the retina.

[0043] FIG15B shows further detailed layering of the retina.

[0044] FIG16 shows a basic general scanning protocol for optical coherence tomography angiography (OCTA).

[0045] 17A-17C show schematic diagrams of pERG, flash ERG, and mfERG, respectively.

[0046] FIG18 shows a schematic diagram of light stimulation of pERG (left side) and a schematic diagram of sample waveforms (right side).

[0047] FIG19A shows a schematic diagram of light stimulation of mfERG (left), a schematic diagram of contact electrodes (middle), and a schematic diagram of the detected electrical waveform (right).

[0048] FIG19B shows a schematic diagram showing the distribution of amplitude in mfERG and the relationship between amplitude and latency represented by a topographic map.

[0049] FIG20 is a schematic diagram showing a comparison of OCT thickness and volume parameters for the young and elderly groups divided by method 1. FIG.

[0050] FIG21 is a schematic diagram showing a comparison of OCT thickness and volume parameters for the young and elderly groups divided by method 2. ...

[0051] FIG22 is a schematic diagram showing a comparison of OCT thickness and volume parameters for the young and elderly groups divided by method 3. FIG.

[0052] FIG23 is a schematic diagram showing a comparison of vascular parameters between the young group and the elderly group. DETAILED DESCRIPTION

[0053] Embodiments of the present invention relate to methods for the integrated analysis of retinal structural and functional biomarkers for identifying CNS cell senescence.

[0054] The terms used herein are only used for the purpose of describing specific embodiments and are not intended to limit the present invention. As used herein, the term "and / or" includes any and all combinations of one or more associated listed items. As used herein, the singular forms "a", "an" and "the" are intended to include plural forms as well as singular forms, unless the context clearly indicates otherwise. It will be further understood that when used in this specification, the terms "include" and / or "comprise" specify the presence of the features, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components and / or groups thereof.

[0055] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this disclosure, and will not be interpreted in an idealized or overly formal sense unless explicitly defined as such herein.

[0056] When the term "about" is used in conjunction with a numerical value herein, it is understood that the value may be within the range of 90% to 110% of the value, i.e., the value may be + / - 10% of the value. For example, "about 1 kg" means 0.90 kg to 1.1 kg.

[0057] The human retina, as an extracranial extension of the brain's nerves and the only visible neurovascular tissue, is an ideal window for monitoring the properties of CNS neurons. As shown in Figure 15A, the human retina can be roughly divided into three different layers: the photoreceptor layer, the ganglion cell layer, and the bipolar cell layer. They can be further divided into more layers, as shown in Figure 15B. Among them, ganglion cells are projection neurons of the vertebrate retina that transmit information from other retinal neurons to other parts of our brain. In order to thoroughly monitor its changes, we need to conduct a comprehensive analysis of the properties of the ganglion cell layer (GCL), including structural, vascular, and functional properties.

[0058] However, current studies have two major limitations. First, most OCT instruments used in current studies cannot separate the GCL and IPL. Second, current studies rarely combine retinal ganglion cell characteristics with retinal vascular and functional characteristics for comprehensive analysis.

[0059] Therefore, in order to solve the above problems, the present application proposes the following method.

[0060] Referring to Figures A-E3 in Figure 1, a comprehensive analysis process of OCT thickness and volume of different layers of the retina in the macular region using three different partitioning methods is shown. A method for providing comprehensive and detailed analysis of OCT thickness and volume is provided by applying three different partitioning methods to five different layers of the retina in the macular region (as shown in Figure 1, B). The method includes: recording an optical tomography angiography (OCT) thickness map using a posterior pole module; applying the Early Treatment Diabetic Retinopathy Study (ETDRS) grid to the OCT thickness map to divide the retina into eight regions (as shown in FIG3); grouping the eight regions according to three different methods as shown in E1, E2, and E3 in FIG1; calculating the average thickness and volume data of each region including the nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), from the nerve fiber layer to the inner plexiform layer (NFL-IPL), and from the ganglion cell layer to the inner plexiform layer (GCL-IPL) (as shown in FIG1B); cropping the OCT thickness map output from the software so that only the right half of the numerical data is retained for OCT thickness and volume analysis (as shown in FIG3); and extracting numerical data from the cropped OCT thickness map by an OCR space text extraction method (as shown in FIG4).

[0061] In addition, the present invention defines a custom parameter named Area, which represents the ratio of the area with a thickness greater than 45 μm on the GCL thickness map to the area of ​​the 6 mm circle of the ETDRS grid. The parameter Area is used as a comprehensive indicator of ganglion cell richness, and combined with the standard ETDRS grid analysis indicators, it helps to more comprehensively evaluate the characteristics and changes of retinal structure. The method for determining Area includes: cropping the OCT thickness map output from the software so that only the left half of the original thermal map image is retained for analyzing Area; extracting areas with a thickness greater than 45 μm and calculating their areas through image processing technology; extracting the area within the 6 mm circle of the ETDRS grid and calculating its area through image processing technology; calculating Area, which is the ratio of the area with a thickness greater than 45 μm on the GCL thickness map to the area within the 6 mm circle of the ETDRS grid. An example of calculating Area is shown in Figure 6.

[0062] In one embodiment, a method for comprehensive and detailed analysis of superficial and deep vascular plexus images using OCTA is provided. The method includes: recording optical coherence tomography angiography (OCTA) vascular image data; extracting a region within a 2 mm circle centered on the macula, as shown in Figure 9, and cropping the rest of the image as a trade-off for retaining more vascular area without losing vascular detail; calculating vascular abundance, vessel length, vessel diameter, and vessel tortuosity by uploading the cropped vascular image to a toolbox called OCTAVA; and calculating foveal avascular zone (FAZ) parameters, including the FAZ area index and the FAZ circularity index. Figure 10 shows an example of image processing and data analysis using OCTAVA.

[0063] In another embodiment, a method for dividing the mfERG report analysis area in two different ways is also provided. The method includes: recording 103 mfERG tests and obtaining mfERG reports; dividing 103 hexagons into four different areas / quadrants with a horizontal line and a vertical line, and calculating the average latency and amplitude of P1 and N1 in the four different quadrants; and dividing the 103 hexagons into six concentric equidistant eccentricities around the central disk, and calculating the average latency and amplitude of P1 and N1 in six different rings. The two division methods are shown in Figures 13A and 13B.

[0064] In another embodiment, a method for integrated analysis of retinal imaging and functional biomarkers is provided. The method comprises applying both the ETDRS grid and 103 hexagons to the retina at the macular region, both centered on the fovea; and correlating the structural and functional changes in the macula region of the retina by correlating the overlapping areas of the two overlays. An image showing two superpositions can be seen in Figure 14. The 1 mm circle of the ETDRS grid corresponds to Ring 1 and Ring 2 of the combination. The 3 mm circle of the ETDRS grid (inner annulus, yellow area in Figure 14) corresponds to Ring 3, and the 6 mm circle of the ETDRS grid (outer annulus, green area in Figure 14) corresponds to Ring 4. The results can be well correlated using this combined analysis.

[0065] The results of retinal imaging analysis and examples of functional changes in the macular region with age are also provided. First, the results revealed trends in thickness changes in the five different layers using the three division methods shown above. To further analyze the results and gain a deeper understanding of the results, the samples were divided into two different age groups: the young group, aged between 22 and 32 years, and the elderly group, aged between 50 and 62 years. An independent sample t-test (t-test) was then performed to assess differences between the two age groups.

[0066] For OCT thickness map analysis, to validate the validity of the defined retinal structural biomarkers, an independent sample t-test was performed between age and structural parameters, including the average thickness / volume of five different layers in different regions as delineated by the three partitioning methods described. For partitioning method 1, NFL thickness in the elderly group was significantly increased in the inner annular zone, GCL thickness was significantly decreased in the outer annular zone and total annular zone, and NFL-IPL thickness was significantly decreased in the outer annular zone. For partitioning method 2, NFL thickness in the elderly group was significantly increased in the total temporal zone, while GCL thickness was significantly decreased in the total superior, total inferior, and total temporal zones. In addition, NFL volume in the elderly group was significantly increased in the total temporal zone, GCL volume was significantly decreased in the total superior and total inferior zones, and NFL-IPL volume was significantly decreased in the total superior zone. For classification method 3, the NFL thickness of the elderly group increased significantly in the inner annular nasal, inner annular temporal, and outer annular temporal regions, the GCL thickness decreased significantly in the inner annular temporal and outer annular inferior regions, the NFL-IPL thickness decreased significantly in the outer annular inferior region, and the GCL-IPL thickness decreased significantly in the outer annular inferior region. Furthermore, the NFL volume of the elderly group increased significantly in the inner annular temporal and outer annular temporal regions, and the GCL volume decreased significantly in the inner annular superior region and outer annular inferior region.

[0067] For OCTA vascular image analysis, to validate the validity of the defined retinal vascular biomarkers, independent sample t-tests were performed between age and vascular parameters, including general parameters such as vessel area density (VAD) and vessel length density (VLD). In addition, the corresponding total / mean / median / variance of vessel length, diameter, and tortuosity were included. For the elderly group, the richness of the vasculature was significantly reduced. Furthermore, the total vessel length was significantly reduced in the elderly group. Furthermore, the mean and median values ​​of vessel diameter and the mean value of vessel tortuosity were significantly increased in the elderly group.

[0068] For mfERG reporter analysis, in order to verify the validity of the defined retinal functional biomarkers, an independent sample t-test was performed between age and functional parameters according to the two different division methods (4-quadrant division and 6-ring division) as described above, including the average amplitude and latency of P1 and N1 in different areas of 103 hexagons. For the elderly group, the latency of P1 was significantly increased in quadrants 3 and 4 (4-quadrant division) and rings 4 and 5 (6-ring division) of the 103 hexagons, and the latency of N1 was also significantly increased in quadrant 1 (4-quadrant division) and ring 3 (6-ring division) of the 103 hexagons.

[0069] Combining the results of OCT thickness map analysis and mfERG report analysis showed that they can be well correlated. In the inner annular zone corresponding to ring 3 (see Figure 14, yellow area), the NFL thickness of the elderly group increased significantly, and the latency of N1 in the elderly group also increased significantly. In the outer annular zone corresponding to ring 4 (see Figure 14, green area), the GCL and NFL-IPL thickness of the elderly group decreased significantly, and the latency of P1 in the elderly group also increased significantly.

[0070] One limitation is that, because the present embodiment of the integrated analysis method differs from conventional analysis modules, it may require very specific data types, including all OCT scan images, OCTA vascular images, and pERG and mfERG test results. In this case, the integrated analysis module cannot adapt to situations where part of the retinal data is missing. Furthermore, the integrated analysis method of the present invention requires that the OCT / OCTA and ERG machines used are capable of properly separating the ganglion cell layer and that the ERG machine can perform 103 mfERG tests.

[0071] Another limitation is the relatively small sample size used to validate the retinal imaging and functional biomarkers in this study: OCT / OCTA scans were performed on 20 subjects in each age group (young and elderly), and pERG and mfERG tests were performed on 10 subjects in each age group. However, significant retinal changes were still observed, particularly in the outer ring of the macular region.

[0072] Regarding the first limitation, although the detection capabilities of the equipment may affect the universality of this analysis module, the goal of this study is to conduct a more comprehensive and detailed analysis of the macular retina from multiple levels, including structure, blood vessels, and function. Therefore, this limitation will not be addressed.

[0073] To overcome the second limitation, more participants could be recruited from staff or students at the University of Hong Kong and the local community to further confirm the validity of this comprehensive analysis module.

[0074] In one embodiment, the present invention provides a comprehensive analysis method for evaluating retinal changes in terms of retinal structure and function. All optical coherence tomography (OCT) and optical coherence angiography (OCTA) images are acquired with a resolution of up to 3 microns, which can provide detailed information at the cellular level. Specifically, the present method uses spectral domain optical coherence tomography (SD-OCT) technology equipped with a vascular imaging module to ensure the accuracy of retinal stratification and vascular system details. With the continuous advancement of computer image processing and analysis technology, in-depth analysis of changes at the retinal cell level can be performed to discover retinal biomarkers related to aging and evaluate the health status of CNS neurons. By integrating retinal structure and function analysis at the cellular level, the method of the present invention can discover reliable retinal biomarkers to provide support for evaluating the health of CNS neurons. This cross-scale integrated analysis method helps to more comprehensively reflect the overall status of the visual system and central nervous system, and provides an important basis for the early diagnosis and treatment of neurodegenerative diseases.

[0075] Commercially available machines with specific designs and detection parameters for image acquisition and reconstruction are used. Changes in RGC morphology and functional levels can be used for early detection of CNS neuronal health.

[0076] First, this method acquires retinal structural biomarkers in the macular region, including detailed analysis of layer thickness and volume data from OCT scans. Next, the retinal vasculature in the macular region is analyzed based on the superficial and deep vascular plexuses obtained from OCTA scans. Next, multifocal electrophysiological monitoring (mfERG) is used to assess retinal function in specific areas. By quantitatively and comprehensively analyzing these three aspects, changes in retinal ganglion cells during aging can be assessed, thereby reflecting the health of CNS neurons.

[0077] This method for comprehensively analyzing retinal ganglion cell properties from a structural, vascular, and functional perspective offers two major advantages. First, this method provides comprehensive analysis of retinal ganglion cell distribution through OCT thickness maps, comprehensive analysis of retinal vasculature through OCTA superficial vascular plexus images, and comprehensive analysis of retinal function through pERG and mfERG reports. This detailed and comprehensive analysis may reveal potential mechanisms underlying CNS neuronal aging. A second advantage is that this comprehensive analysis method analyzes ganglion cell layer thickness alone, without combining it with the inner plexiform layer, thereby enabling analysis of ganglion cell properties at the cellular level.

[0078] Using the properties of human retinal ganglion cells, which are projection neurons of the vertebrate retina, to monitor CNS neuronal health is a new and previously underexplored approach.

[0079] The present invention has customized retinal structural parameters, which have been shown to be effective for analyzing retinal ganglion cell characteristics and assessing CNS neuronal health. In addition to detailed analysis of retinal vascular and functional changes, the present invention also includes a comprehensive analysis module. This module can effectively promote the discovery of retinal biomarkers and provide support for reliable monitoring of CNS status. By integrating quantitative analysis of multiple aspects such as retinal structure, blood vessels, and function, the present invention constructs a comprehensive evaluation framework. This cross-scale integrated approach helps to more accurately reflect the overall status of the visual system and central nervous system, and provide support for early diagnosis and treatment of neurological diseases.

[0080] Embodiments of the present invention focus on cellular level analysis and can correlate retinal structural and functional changes to reveal the underlying mechanisms of CNS cell aging.

[0081] In one embodiment, the present invention provides a system and method for a comprehensive analysis module of retinal structural and functional biomarkers in great detail. Retinal biomarkers can facilitate early detection of central nervous system neuronal health by studying retinal neuronal characteristics at the cellular level based on OCT thickness maps, OCTA vascular images, and mfERG reports. This allows for timely intervention to mitigate neuronal progression and maintain neuronal health in the CNS.

[0082] The quantitative structural and functional biomarkers of the retina generated by the present invention can be used to develop intervention plans related to or potentially related to retinal aging and central nervous system (CNS) neuronal decline. These quantitative indicators help to achieve early detection and timely intervention treatment. In addition, the system and method of the present invention can also be used as a routine clinical examination or community screening tool based on retinal data to monitor the overall health of CNS neurons. This non-invasive assessment method based on the retina provides a convenient and efficient solution for clinicians and researchers, which is expected to promote the early detection and management of neurological diseases.

[0083] In a preferred embodiment, the system and method define and quantify retinal imaging and functional parameters. Participants were divided into two age groups. The young group was aged 22 to 32 years, and the elderly group was aged 50 to 62 years. Inclusion criteria for the retina were (1) best corrected visual acuity of 20 / 50 or better; (2) intraocular pressure less than 21 mmHg. Exclusion criteria were (1) presence of significant refractive error, intraocular pressure >21 mmHg; (2) media opacity; (3) concomitant ocular disease; (4) history of glaucoma or retinal pathology or systemic conditions that may affect the visual system; (5) history or evidence of ocular or neurological disease of any nature; (6) best corrected visual acuity ≤20 / 30 (±0.2, in Log Mar) based on the Snellen scale. All subjects were required to undergo a complete ophthalmological evaluation, including pupillary, anterior segment and funduscopy.

[0084] OCT thickness map analysis

[0085] In some embodiments, the present invention uses the posterior pole module to obtain OCT thickness maps of different layers. For example, in some embodiments of the system of the present invention, the OCT images of the participant include, but are not limited to, thickness maps of the nerve fiber layer (NFL), ganglion cell layer (GCL), and inner plexiform layer (IPL). All thickness maps were obtained using OCT (SD-OCT) built-in software using two different applications: Fast Macular Cube (for macular analysis) and Retinal Nerve Fiber Layer (RNFL) Glaucoma (for peripapillary assessment). The OCT thickness maps output by the software shown in Figure 3 were then cropped accordingly. For thickness and volume analysis, only the right half was required for numerical data extraction. For custom parameter area analysis, only the left half was required for image processing.

[0086] In some embodiments, for detailed analysis, an Early Treatment Diabetic Retinopathy Study (ETDRS) grid comprising circles having diameters of 1 mm, 3 mm, and 6 mm, respectively, is overlaid on the OCT thickness map to divide the OCT thickness map into nine regions (or zones), as shown in FIG2 , with the following labels:

[0087] 1. C1: foveal area;

[0088] 2. S3, N3, I3, T3: the inner ring superior, inner ring nasal, inner ring inferior, and inner ring temporal regions of the inner ring area between 1 mm and 3 mm, respectively;

[0089] 3. S6, N6, I6, T6: the outer ring superior, outer ring nasal, outer ring inferior and outer ring temporal areas between 3mm and 6mm, respectively.

[0090] The Early Treatment Diabetic Retinopathy Study (ETDRS) analysis method allows for the extraction of valuable quantitative data from retinal thickness maps obtained using optical coherence tomography (OCT) imaging. The ETDRS grid is a standardized, reliable, and reproducible method for analyzing and reporting retinal thickness, facilitating better comparison of results across research and clinical settings. Overall, ETDRS charts are considered more accurate and reproducible for analysis and comparison. This is because the ETDRS grid corresponds to anatomical and functional regions of the retina, providing clinically relevant information on retinal structure and its potential impact on retinal function. Based on the ETDRS grid, volume and thickness data for different retinal layers (such as the ganglion cell layer (GCL) and its adjacent layers (NFL and IPL)) can be extracted and analyzed across different subfields. These quantitative metrics facilitate a more comprehensive assessment of retinal structural characteristics, providing valuable insights for the diagnosis and monitoring of various retinal diseases.

[0091] Figure 3 shows an example of a thickness map of the ganglion cell layer and the statistics generated by the software. Different colors on the thickness map represent different thickness values ​​for that region. The lighter the color, the thicker the area. For each of the nine regions, the built-in analysis software provides the average thickness and volume information for that specific region.

[0092] In some embodiments, the present invention analyzes and quantifies thickness maps of the ganglion cell layer (GCL) and its adjacent layers, including the nerve fiber layer (NFL) and inner plexiform layer (IPL), to identify signs of neurodegeneration. The average thickness and volume of the NFL-IPL from the nerve fiber layer to the inner plexiform layer and the GCL-IPL from the ganglion cell layer to the inner plexiform layer are also analyzed to gain a deeper understanding.

[0093] In a preferred embodiment, the present invention quantifies retinal imaging parameters including but not limited to the average thickness and volume of different regions of the ETDRS (Early Treatment Diabetic Retinopathy Study) grid of different layers of the NFL, GCL, IPL, GCL-IPL and NFL-IPL.

[0094] In some embodiments, the present invention extracts useful data from the output images of the SD-OCT machine using OCR Space, an online image processing method that recognizes and extracts numbers or strings from images and converts them into a parseable format, such as text or JSON (Badwaik, Mahmood, & Raza, 2017). Figure 4 shows an example of OCR Space input and output data, which correctly extracts all useful data, including the average thickness and volume of nine regions.

[0095] In a specific embodiment, the present invention divides the 1, 3, and 6 mm ETDRS grids into three distinct zones, including the fovea, i.e., the area within the 1 mm circle; the inner annulus, i.e., the area between the 1 mm and 3 mm circles; and the outer annulus, i.e., the area between the 3 mm and 6 mm circles. An illustration of this division is shown in FIG5A . Furthermore, the total annulus is the combined area of ​​the inner annulus and the outer annulus. For each of the inner annulus, outer annulus, and total annulus zones, mean thickness and volume data can be obtained and then analyzed using a two-tailed independent sample t-test to calculate a significance value (P) between the two age groups.

[0096] In a specific embodiment, the present invention divides the 1, 3, and 6 mm ETDRS grids into four distinct regions: total superior (S3 and S6), total nasal (N3 and N6), total inferior (I3 and I6), and total temporal (T3 and T6). An illustration of the division is shown in FIG5B. For each of these four distinct regions, volume and thickness data can be obtained and then analyzed using a two-tailed independent sample t-test to calculate a significance value (P) between the two age groups.

[0097] In a specific embodiment, the present invention divides the 1, 3, and 6 mm ETDRS grids into eight distinct regions, including the inner annular region between 1 mm and 3 mm: superior inner annular region (IS), nasal inner annular region (IN), inferior inner annular region (II), and temporal inner annular region (IT); and superior outer annular region (OS), nasal outer annular region (ON), inferior outer annular region (OI), and temporal outer annular region (OT). An illustration of the division is shown in FIG5C . For each of these eight distinct regions, volume and thickness data can be obtained and then analyzed using a two-tailed independent sample t-test to calculate a significance value (P) between the two age groups.

[0098] In some embodiments, the present invention defines a parameter called Area, which represents the ratio of the area with a thickness greater than 45 μm on the GCL thickness map to the area of ​​a 6 mm circle of the ETDRS grid. This indicator is based on the fact that the average size of human retinal ganglion cells is about 10 μm, and it is believed that areas greater than 45 μm represent sufficient distribution of ganglion cells in the macular area. As shown in Figure 6, the calculation process of this parameter includes image processing and extraction of areas with a thickness greater than 45 μm using Python OpenCV and Shapely packages. The relevant code can be found in Example 8. The "Area" parameter is regarded as a comprehensive indicator that can reflect the richness of ganglion cells. Since the analysis process only focuses on a certain area of ​​the ETDRS grid, the integrity of the analysis may be lost. Therefore, the introduction of the "Area" parameter can make up for this deficiency and provide a supplementary indicator for effectively evaluating the status of ganglion cells.

[0099] OCTA vascular image analysis

[0100] In some embodiments, the methods of the present invention quantify retinal microvascular parameters, including but not limited to vessel area abundance, vessel diameter, vessel length, vessel tortuosity, vessel branch point abundance, and vessel fractal dimension. Figure 7 shows all vessel characteristics, and the present invention calculates the mean, median, variance, skewness, and kurtosis of the distribution of these characteristics. In a further preferred embodiment, the methods quantify parameters of the foveal avascular zone (FAZ), including but not limited to FAZ area, FAZ perimeter, and FAZ circularity.

[0101] In certain embodiments, the present invention acquires OCTA image data. Optical coherence tomography angiography, also known as OCTA, is a relatively new non-invasive method for visualizing retinal microvasculature. The basic general scanning protocol for optical coherence tomography angiography (OCTA) is illustrated in Figure 16. (a) Repeated B-scans are performed on the "x" fast axis at each "y" slow scan axis point to detect relative flow signals. (b) is a top view of the same general scanning pattern as (a), with repeated B-scans performed on the "x" fast axis at each "y" position along the slow axis. OCTA rapidly acquires multiple cross-sectional images from a single location and uses motion contrast to distinguish between moving and static elements. Using this technique, microvasculature of the superficial and deep capillary plexuses can be separately visualized and analyzed. For example, in some embodiments of the systems and methods of the present invention, participant OCTA images (including, but not limited to, two consecutive 3 mm x 3 mm coronal OCTA images of the superficial and deep vascular plexuses) are generated, for example, by SD-OCT with an angiography module. Figure 8 shows exemplary acquired superficial and deep vascular plexus images. The present invention extracts the region of interest (ROI) within a 2mm circle of the superficial vascular plexus image centered on the macula. This is a trade-off between including more vascular area while preserving vascular detail. Figure 9 shows the original image and the processed image, which extracts the region within a 2mm diameter circle centered on the fovea.

[0102] In some embodiments, the present invention utilizes an open source toolbox called OCTAVA, which is developed using MATLAB and based on the image processing functions in ImageJ (Untracht et al., 2021). The toolbox can be used to perform detailed quantitative analysis of optical coherence tomography angiography (OCTA) images. As shown in Figure 10, after uploading the OCTA image, the toolbox will first binarize and skeletonize the original image. Then, based on the extracted vascular structure, the toolbox can analyze various vascular parameters, such as vascular abundance, diameter, length, and tortuosity, and output their distribution data to an Excel file. Next, the two-tailed independent sample t-test in SPSS software is used to perform statistical analysis on the vascular parameters between different age groups, and the significance value (P) is calculated.

[0103] In some embodiments, the present method uses a method defined by a patent of the University of Hong Kong (Chiu, 2020) to calculate the FAZ parameter in the vascular image. The FAZ is defined as the area within the fovea centralis in the center of the human retina that is devoid of retinal blood vessels (Shiihara et al., 2017).

[0104] In some embodiments, the systems and methods calculate the largest polygon in the avascular fovea, i.e., a polygon that does not cover any positive vascular signal in the resulting image. In some embodiments, the designation (0, 0) corresponds to the upper left corner of the vascular image, and the center of the foveal region is (x inner ,y inner ), and the side length of the blood vessel image is "s". In some embodiments, Θ = 0, 3, 6, 9, ..., 357, that is, considering the inner ,y inner ) to (x iner +s·sin sinθ,y inner +s·cos cosθ) for a total of 120 θ pixels of all pixels within the image on a directed line segment. In some embodiments, the first pixel with a positive signal, i.e., a portion of a blood vessel, is recorded as one of the endpoints of the FAZ. If the value of all pixels within the image on the directed line segment is zero, then that particular θ pixel is skipped. If fewer than three searches return a positive signal, an error is reported. In certain embodiments, a polygon is defined by connecting all n=120 endpoints in the order in which they are found in the approximate FAZ. In other embodiments, when fewer than all endpoints are used (some of which searches do not find any positive pixels), a polygon is defined by connecting fewer than 120 endpoints to divide the FAZ.

[0105] In some embodiments, the systems and methods of the present invention calculate the FAZ Area Index (FAI). In some embodiments, the systems and methods divide the vertices of the FAZ polygon into (a1, b1), (a2, b2), ..., ... n ,b n In certain embodiments, the systems and methods use the shoelace formula to calculate the area of ​​a polygon:

[0106] In some embodiments, the systems and methods define FAI as the FAZ polygon area divided by the pixel area of ​​the vessel image. Furthermore, if the vessel image is always focused on a 3×3 area, the FAI is comparable between images of different sizes.

[0107] In some embodiments, the systems and methods of the present invention calculate the FAZ perimeter index (FPI). In some embodiments, the systems and methods calculate the perimeter of the FAZ polygon using the following formula:

[0108] In some embodiments, the FPI is defined as the FAZ polygon perimeter divided by the side length of the vessel image (in pixels). Furthermore, if the vessel image is always focused on a 3×3 area, the FPI is comparable between images of different sizes.

[0109] In some embodiments, the systems and methods of the present invention calculate the FAZ roundness index (FACI). In a specific embodiment, the FACI quantifies the irregularity of the shape of the divided FAZ and is calculated by the following formula:

[0110] When FACI is equal to 1, the FAZ is a perfect circle. In some embodiments, when the FACI value is greater than 1, the FAZ has a more irregular shape.

[0111] ERG test report analysis

[0112] In some embodiments, the present invention will quantify retinal function parameters including, but not limited to, the latencies of N35, P50, and N95 as measured in pERG examinations, and the amplitudes of N35 to P50 and P50 to N95. In addition, parameters include the amplitudes of P1 and N1 as measured in mfERG examinations, and the latencies of P1 and N1.

[0113] The electroretinogram (ERG) is a diagnostic test that measures the electrical activity of the retina in response to light stimulation. As shown in Figures 17A, 17B, and 17C, there are three main types of ERG examinations: full-field pattern ERG, flash ERG, and mfERG. Using different types of ERG tests to determine the functional status of the retina allows clinicians and researchers to assess patients for retinal conditions, reliably monitor retinal function over time, and evaluate the efficacy of retinal treatments.

[0114] In a pattern ERG (pERG), the light stimulus is an alternating checkerboard pattern, which evokes an electrical response from the ganglion cells that make up the inner retina. The pERG is commonly used to detect diseases that affect ganglion cells, as well as many neurodegenerative diseases.

[0115] Specifically, pattern ERG is a non-invasive technique used to assess the function of the inner retina, particularly retinal ganglion cells. It activates retinal ganglion cells through reverse checkerboard stimulation, generating electrical responses that can be detected using electrodes placed in the eye or surrounding area.

[0116] The macula is primarily stimulated. As shown in the sample waveform of Figure 18, the initial negative response of retinal ganglion cells to the onset of stimulation N35 occurs at 35 milliseconds. Furthermore, P50 is the primary parameter used for analysis in graphical ERG analysis. It is the positive peak component and reflects the activity of the distal ganglion cells. N95 is the negative peak component and reflects the activity of the inner retinal layers. Graphical ERG is most effective in distinguishing abnormalities at the ganglion cell level from distal lesions. The method of the present application extracts and analyzes the latencies of N35, P50, and N95, as well as the amplitudes of N35-P50 and P50-N95.

[0117] In a full-field ERG (ffERG), electrodes record the total electrical response of retinal cells. The total electrical response is a mix of results from a range of different cells and areas in the retina, meaning that the full-field ERG (ffERG) shows the state of the retina as a whole.

[0118] In mfERG (mfERG), a light stimulus displays a series of hexagons that flash between white and black in a pseudorandom sequence. mfERG is best suited to measure dysfunction in the central retinal area.

[0119] Specifically, the mfERG test shows visual function across the macula. The test divides the macula into multiple hexagonal areas. The stimulus is a random sequence of black and white hexagons, alternating multiple times per second. The picture shown on the left side of Figure 19A is an example of a series of black and white hexagons. These stimuli will then be received by the human eye, and the contact electrodes placed in the eye will record the electrical activity in response to the stimuli within each hexagonal area. Finally, the results are shown in the picture on the right, and for each hexagonal area, the response is displayed in the form of an electrical waveform. The built-in software of the mfERG machine can also use the concept of topographic maps to provide a 3D map of the distribution of important measurements of the indication waveform, as shown in Figure 19B. This application extracted and analyzed the latency and amplitude of N1 and P1.

[0120] In a specific embodiment, the present invention obtains pERG and mfERG test results for analysis because the present invention focuses on the macular area and mfERG divides the central retinal area into 103 hexagons, which provides a global assessment of retinal electrical activity.

[0121] For example, in some embodiments of the comprehensive analysis of the present invention, the pERG and mfERG test results of the participant include, but are not limited to, using, for example, a RETI-port / scan 21ERG machine to generate a graphic ERG report showing an electrical activity waveform with N35, P50, N95 latencies and P50 and N95 amplitudes, and an mfERG report showing the overall retinal electrical activity waveforms of different areas of the retina. For accuracy, the graphic ERG 48-minute test is repeated twice for both eyes without dilating the pupil. And for the mfERG test, both eyes are examined separately, and the resolution level is set to 103 segments, and two 50Hz smoothing filters are applied. Corneal contact electrodes (BA / ERG-Jet) and gold foil electrodes are necessary for both graphic ERG and mfERG tests. In other embodiments, a test report of the output from the ERG machine is generated.

[0122] In some embodiments, the present invention extracts useful data from the output ERG report using OCR Space, an online image processing method that can recognize and extract numbers or strings from images and convert them into a parseable format, such as text or JSON (Badwaik, Mahmood, & Raza, 2017). Figure 11 shows an example use of OCR Space, in which an mfERG report is uploaded and all useful data, including the amplitude and latency of P1 and N1 within each ring, are extracted.

[0123] In some embodiments, the present invention extracts the N35, P50, and N95 latencies, as well as the P50 and N95 amplitudes, from the pERG 48-minute test report. An example of the output of the pERG report is shown in FIG12 . To ensure accuracy, the pERG test was repeated twice, and the test data for P-ERG1 and P-ERG2 were recorded and analyzed. For each parameter, a two-tailed independent sample t-test was used to analyze the differences between the groups and calculate the significance value (P) between the two age groups.

[0124] In some embodiments, the system and method of the present invention divides 103 hexagons into 4 quadrants, as shown in Figure 13A. For each quadrant, the average electrical waveform is generated by the built-in software, and the amplitudes of P1 and N1 and the latency of P1 and N1 will be generated and displayed in the mfERG test report. The first column Amp.P1 has a unit different from that of the other columns, which is nV divided by the square of the degree. Its value represents the amplitude of P1 of a specific hexagon divided by the area of ​​the hexagonal area to achieve the purpose of standardization. The present invention will use a two-tailed independent sample t test to analyze the differences between the groups and calculate the significance value (P) between the two age groups for each parameter.

[0125] In some embodiments, the systems and methods of the present invention divide the 103 hexagons into 6 concentrically equidistant eccentrics around the central disk, as shown in Figure 13B. The range of each ring is shown as: Ring 1, from 0° to 1°; Ring, from 1° to 4°; Ring 3, from 4° to 8°; Ring 4, from 8° to 12°; Ring 5, from 12° to 17°; Ring 6, from 17° to 21°. For each ring, the average electrical waveform is generated by the built-in software, and the amplitudes of P1 and N1 and the latencies of P1 and N1 will be generated and displayed in the mfERG test report. For each parameter, the present invention uses a two-tailed independent sample t-test to analyze the differences between the groups and calculate the significance value (P) between the two age groups.

[0126] Comprehensive analysis and application

[0127] In some embodiments, the systems and methods of the present invention overlay 1, 3, and 6 mm ETDRS grids onto the 103 hexagons of the mfERG test report and associate them for comprehensive analysis. Two overlays with the same center at the macula on the same retinal image are shown in Figure 14. The fovea area is within the 1 mm circular area and corresponds to ring 1 and ring 2 of the mfERG result (from 0 ° to 4 °, the red part in Figure 14). The inner ring area between 1 mm and 3 mm corresponds to ring 3 of the mfERG result (from 4 ° to 8 °, the yellow part in Figure 14). The outer ring area between 3 mm and 6 mm corresponds to ring 4 of the mfERG result (from 8 ° to 12 °, the green part in Figure 14).

[0128] In certain embodiments, the methods of the present invention correlate retinal structural and functional changes through comprehensive and detailed analysis of OCT thickness maps, OCTA vascular images, pERG, and mfERG test reports. Correspondence is established using 1, 3, and 6 mm ETDRS grids and 103 hexagonal grids centered on the fovea.

[0129] In some embodiments, the method of the present invention can collect a user's retinal images and retinal function report and output a comprehensive assessment of retinal structure, vascularity, and function. Parameters from these three aspects can be correlated using a 1, 3, and 6 mm ETDRS grid and 103 hexagonal divisions with the same center at the macula, as shown in FIG14 .

[0130] In some embodiments, the comprehensive analysis module of the present invention can be incorporated into SD-OCT clinical equipment, enabling clinicians to more conveniently utilize the module and enhance the depth of retinal image analysis. In a preferred embodiment, the comprehensive analysis module can directly quantify retinal structural and vascular parameters. Once the clinic completes retinal imaging, the module outputs quantitative results without the need for tedious image processing and quantification, facilitating image evaluation and analysis by clinicians.

[0131] This invention provides a comprehensive analysis system and method for imaging and functional biomarkers of aging retina. This method comprehensively analyzes the structural, vascular, and functional properties of retinal ganglion cells by analyzing three different data formats: ganglion cell layer thickness maps obtained by optical coherence tomography (OCT), superficial vascular information obtained by OCTA, and functional information obtained by mfERG testing. This multimodal data analysis not only analyzes ganglion cell properties at the cellular level but also provides noninvasive insights into the health of central nervous system neurons and the mechanisms of aging.

[0132] In addition, the present invention also provides examples of analyzing retinal imaging and functional changes in aging to facilitate the identification of retinal imaging and functional biomarkers, enabling effective, quantitative and reliable evaluation, identification, assessment and intervention recommendations for the aging process.

[0133] Materials and methods

[0134] The following are examples illustrating procedures for practicing the present invention. These examples should not be construed as limiting. Unless otherwise indicated, all percentages are by weight and all solvent mixture ratios are by volume.

[0135] Example 1 - Method for quantitative OCT image analysis of ganglion cell distribution

[0136] The present invention uses the posterior pole module of SD-OCT to obtain OCT thickness maps of five different layers, including the NFL, GCL, IPL, NFL-IPL, and GCL-IPL. The ETDRS grid is then applied to process and analyze the OCT thickness maps. ETDRS includes circles with diameters of 1 mm, 3 mm, and 6 mm, respectively, overlaid on the OCT thickness map to divide the map into nine regions, as shown in Figure 2. C1 is the foveal region; S3, N3, I3, and T3 are the inner superior, inner nasal, inner inferior, and inner temporal regions of the inner annular region between 1 mm and 3 mm, respectively; and S6, N6, I6, and T6 are the outer superior, outer nasal, outer inferior, and outer temporal regions of the outer annular region between 3 mm and 6 mm, respectively. The present invention crops the original thickness map report output by the software to retain only the right half of the numerical data, as shown in Figures 3 and 4. This allows the OCR Space text extraction tool to correctly identify the average thickness and volume data of the nine regions on the image, extract them, and convert them into an Excel file for data analysis. The extracted data were then comprehensively analyzed using three different methods.

[0137] The first analysis method divided the macular area into three zones: the inner circle, defined as the foveal area within 1 mm; the inner annulus, defined as the annular area between 1 and 3 mm; and the outer annulus, defined as the annular area between 3 and 6 mm. Furthermore, the total annulus was defined as the sum of the inner and outer annular areas (TA). Based on this partitioning scheme, the mean thickness and volume data of the IA, OA, and TA zones were extracted and analyzed for five different layers.

[0138] The second analysis method divided the macular area into four zones: S, N, I, and T, representing the superior, nasal, inferior, and temporal regions of an annular area ranging from 1 mm to 6 mm. Based on this partitioning, the average thickness and volume data of four zones in five different layers were extracted and analyzed.

[0139] The third analysis method divided the macula into eight zones: IS, IN, II, IT (inner annular zone (1 mm to 3 mm) with inner annular superior, inner annular nasal, inner annular inferior, and inner annular temporal regions); and OS, ON, OI, OT (outer annular zone (3 mm to 6 mm) with outer annular superior, outer annular nasal, outer annular inferior, and outer annular temporal regions). This analysis method extracted and analyzed the average thickness and volume data of the eight zones in five different layers.

[0140] The present invention defines a custom parameter called Area. The Area parameter represents the ratio of areas with a GCL thickness greater than 45 μm to a 6-mm circle on the ETDRS, serving as a comprehensive indicator of ganglion cell richness. Figure 6 illustrates the process of extracting areas with a thickness greater than 45 μm and calculating the Area ratio. Using the Area parameter as a comprehensive indicator of ganglion cell richness provides valuable supplementary information for retinal structural analysis.

[0141] Example 2 - Method for quantitative OCTA image analysis of vascular properties

[0142] The present invention acquired 3×3 mm coronal OCTA images of the superficial vascular plexus (SVP) and deep vascular plexus (DVP) from an SD-OCT machine with an angiography module. As a trade-off strategy, the researchers selected regions of interest (ROIs) with the following characteristics:

[0143] 1. The vascular area is larger, that is, it has a wider vascular coverage area.

[0144] 2. No loss of vascular details, that is, the detailed characteristics of the blood vessels can be retained without being oversimplified or blurred due to image processing.

[0145] Figure 9 shows the ROI determined using this selection principle: a 2mm circle in the superficial vascular plexus image centered on the macula. By balancing vascular area and detail preservation, this ROI facilitates a more comprehensive analysis of vascular structural features in OCTA images.

[0146] Next, all vascular parameters, including the number of nodes, branch point abundance, fractal abundance, and vessel length, diameter, and tortuosity, were obtained by uploading all SVP and DVP images cropped into 2 mm circles to the OCTAVA toolbox, an OCTA vascular image analysis software developed based on ImageJ functions ( Untracht et al., 2021 ).

[0147] To further quantify the FAZ characteristics, the FAZ area index (FAI), FAZ perimeter index (FPI), and FAZ circularity index (FACI) were calculated based on a method provided by another Hong Kong University patent (Chiu, 2020). This method divides the vertices of the FAZ polygon into (a1, b1), (a2, b2), ..., (a n ,b n ). This procedure calculates the area of ​​a polygon using the shoelace formula:

[0148] As explained, the process also defines FAI as the FAZ polygon area divided by the pixel area of ​​the vessel image. FAI is comparable between images of different sizes, and the vessel image is always focused on a 3×3 area.

[0149] The procedure calculates the FAZ Perimeter Index (FPI) using the following formula:

[0150] The procedure defines FPI as the perimeter of the FAZ polygon divided by the side length of the vessel image (in pixels). The FPI is comparable between images of different sizes, and the vessel image is always focused on a 3×3 area.

[0151] The present invention also calculates the FAZ roundness index (FACI). FACI is used to quantify the irregularity of the shape of the divided FAZ and is calculated by the following formula:

[0152] When FACI is equal to 1, the FAZ is a perfect circle, and when the FACI value is greater than 1, the FAZ has a more irregular shape.

[0153] Example 3 - Methods for quantitative ERG reporter analysis of retinal function

[0154] The present invention uses a Reti-Port / scan 21 ERG machine to obtain pERG and mfERG reports. To ensure accuracy, a 48-minute pERG test is repeated twice for each eye without dilating the pupil, and an mfERG test is also performed for each eye. The mfERG test resolution is 103 hexagons, and two 50Hz smoothing filters are applied. During the test process, corneal contact electrodes (BA / ERG-Jet) and gold foil electrodes are used for both pERG and mfERG tests.

[0155] The present invention uses the text extraction method provided by OCR Space to organize all useful ERG test data and prepare them for subsequent analysis and comparison. The text extraction method provided by OCR Space is an online image processing method that can recognize and extract numbers or strings from images and convert them into analyzable formats, such as text or JSON. Figure 11 shows an exemplary use of OCR Space, including uploading an mfERG report and extracting all useful data (amplitude and latency of P1 and N1 within 6 rings and amplitude and latency of P1 and N1 in 4 quadrants). The present invention also extracts the latency of N35, P50 and N95, as well as the amplitude of P50 and N95 from the pERG 48-minute test report. Through the automated OCR Space text extraction method, all useful ERG test data are organized and prepared for subsequent analysis and comparison.

[0156] Example 4 - Method for integrated analysis of both retinal imaging and functional biomarkers

[0157] The present invention associates retinal structure with functional changes through a comprehensive analysis of OCT thickness maps, OCTA vascular images, pERG and mfERG test reports. With the macula as the center, 1, 3, and 6 mm ETDRS grids are used to spatially align OCT and ERG data to establish a corresponding relationship. This process superimposes 1, 3, and 6 mm ETDRS grids on the 103 hexagons of the mfERG test report and spatially aligns the results for comprehensive analysis. The foveal area is within the 1 mm circular area and corresponds to rings 1 and 2 (0° to 4°) of the mfERG results. The inner ring zone of the inner circle area between the 1 mm and 3 mm circles corresponds to ring 3 (from 4° to 8°) of the mfERG results. The outer ring zone of the outer circle area between the 3 mm and 6 mm circles corresponds to ring 4 (from 8° to 12°) of the mfERG results. Figure 14 is a schematic diagram of two superimposed images on the same macular area.

[0158] Example 5 - Measurement of Retinal Imaging and Functional Biomarkers with Integrated Analysis Module

[0159] By comprehensively utilizing the above methods, the present invention has identified a series of reliable retinal structural and functional biomarkers. Structural parameters include the average thickness and volume of different regions of five different layers (NFL, GCL, IPL, NFL-IPL, GCL-IPL); vascular parameters include the number of vascular nodes, vascular area abundance, branch point abundance, fractal abundance, as well as vascular length, vascular diameter, and vascular tortuosity; functional parameters include the P50 and N95 latency and amplitude in the pERG test report, and the P1 and N1 latency and amplitude in different regions in the mfERG test report. To verify the effectiveness of these biomarkers, the present invention conducted a comprehensive analysis of 42 OCT thickness maps, 42 OCTA vascular images, 44 pERG reports, and 20 mfERG reports from 42 healthy subjects. All OCT thickness maps were obtained from OCT structural scans of the macular area using SD-OCT equipped with an angiography module with an 815nm IR laser and a 30-degree lens. All OCT superficial / deep vascular plexuses used were from OCTA scans performed across the macular area within a range of 10°×10° (3mm×3mm). In addition, all ERG tests including pERG tests and mfERG tests were performed using RETI-port / scan 21. The method of the present invention was used to detect different biomarkers representing the overall retinal neuronal properties at the macular area. The participants were divided into two age groups. The young group was 22 to 32 years old, including data from 22 eyes of 11 participants, with a mean age of 25.45 and a standard deviation of 3.52. The elderly group was 50 to 62 years old, including 15 eyes from 9 participants, with a mean age of 54.20 and a standard deviation of 3.91. Summary of the Invention Summary of the Invention Statistics were generated for different retinal biomarkers (for each subject's eye, thickness maps of different layers (NFL, GCL, IPL), SVP images, and pERG and mfERG reports were collected), two-tailed independent sample t-tests were performed between the two age groups, and Pearson correlation tests with 95% confidence interval level were performed, and two-tailed p-values ​​were calculated.

[0160] The results of comprehensive analysis using the systems and methods of the present invention can be seen in the following examples. In Example 6, many of the defined parameters were significantly correlated with age and can be used as biomarkers to assess changes in retinal characteristics with human age and to analyze CNS neuronal health. As shown in Example 7, comprehensive analysis can also provide insights into the underlying biological mechanisms of retinal aging changes.

[0161] Example 6 - Computational Retinal Imaging and Functional Biomarkers

[0162] In order to evaluate the validity of the defined retinal structure and parameters, the present invention uses SPSS to carry out two-tailed independent sample t-test to evaluate the changes in retinal structure and function parameters between the two age groups. For OCT thickness map analysis, Pearson bivariate correlation test is carried out between the average thickness and volume data of age group and NFL, GCL, IPL, NFL-IPL, GCL-IPL respectively. For detailed analysis, the present invention covers the ETDRS grid with a diameter of 1mm, 3mm and 6mm on the OCT thickness map and divides it into 9 areas (as shown in Figure 2). For the thickness map of 5 different layers, the inventive method provides 3 kinds of division methods (reference example 1) for the area in the ETDRS grid. The analysis results are shown in the table below. Asterisk (*) represents the significant difference between the two age groups. One star (*) represents p<0.05, and two stars (**) represent p<0.01.

[0163] The analysis results of the changes in OCT thickness and volume parameters are shown in Figures 20, 21, and 22. The present invention uses three different retinal area division methods to compare the differences in structural parameters between the elderly group and the young group. When using division method 1, it can be seen from Figure 20 that the NFL thickness of the elderly group increased significantly in the inner annular zone area; the GCL thickness of the elderly group decreased significantly in the outer annular zone area and the total annular zone area; the NFL-IPL thickness of the elderly group decreased significantly in the outer annular zone area. When using division method 2, it can be seen from Figure 21 that the NFL thickness of the elderly group increased significantly in the total temporal area; the GCL thickness of the elderly group decreased significantly in the total superior area, the total inferior area, and the total temporal area; the NFL volume of the elderly group increased significantly in the total temporal area; the GCL volume of the elderly group decreased significantly in the total superior and total inferior areas; and the NFL-IPL volume of the elderly group decreased significantly in the total superior area. When using division method 3, it can be seen from Figure 22 that the NFL thickness of the elderly group increased significantly in the inner annular nasal, inner annular temporal and outer annular temporal regions; the GCL thickness of the elderly group decreased significantly in the inner annular temporal and outer annular inferior regions; the NFL-IPL thickness of the elderly group decreased significantly in the outer annular inferior region; the GCL-IPL thickness of the elderly group decreased significantly in the outer annular inferior region; the NFL volume of the elderly group increased significantly in the inner annular temporal region and the outer annular temporal region; and the GCL volume of the elderly group decreased significantly in the inner annular superior region and the outer annular inferior region.

[0164] The analysis results of the changes in OCTA vascular parameters are shown in Figure 23. In order to verify the validity of the defined retinal vascular parameters, the present invention used an independent sample t-test to compare the differences in vascular parameters between different age groups. The vascular parameters analyzed included general parameters such as vascular area abundance (VAD) and vascular length abundance (VLD). In addition, the corresponding total / mean / median / variance of vascular length, diameter and tortuosity were included. As can be seen from the figure, VAD and VLD, which represent vascular abundance, were significantly reduced in the elderly group. In addition, the total vascular length of the elderly group was significantly reduced, while the mean and median of vascular diameter and the mean of vascular tortuosity were significantly increased.

[0165] The analysis results of the changes in retinal function parameters are shown in Tables 1A to 1C. In order to verify the validity of the defined retinal function parameters, the present invention uses an independent sample t-test to compare the differences in functional parameters between different age groups. The functional parameters analyzed include the average amplitudes and latencies of P1 and N1 in different areas obtained according to the two different partitioning methods described in Example 3 (4-quadrant partitioning method and 6-ring partitioning method). It can be seen that the latency of P1 in the elderly group increased significantly in quadrants 3 and 4 (4-quadrant partitioning method) and rings 4 and 5 (6-ring partitioning method), and the latency of N1 also increased significantly in quadrant 1 (4-quadrant partitioning method) and ring 3 (6-ring partitioning method).

[0166] Table 1A Mean differences in functional parameters (pERG) between different age groups

[0167] Table 1B Mean differences in functional parameters (mfERG 4 quadrants) between different age groups

[0168] Table 1C Mean differences in functional parameters (mfERG 6 rings) between different age groups

[0169] The results of the combined OCT thickness map analysis and mfERG report analysis showed that they had good correlation. In the inner annular zone corresponding to ring 3 (see Figure 14, yellow area), the NFL thickness of the elderly group was significantly increased, and the N1 latency of the elderly group was also significantly increased. In the outer annular zone corresponding to ring 4 (see Figure 14, green area), the GCL and NFL-IPL thickness of the elderly group were significantly reduced, and the P1 latency of the elderly group was also significantly increased.

[0170] The present invention also performed Pearson correlation tests on retinal structural and functional parameters and age, and calculated two-tailed p-values ​​with 95% confidence intervals using SPSS. Retinal structural parameters included the average thickness and volume of the NFL, GCL, IPL, NFL-IPL, and GCL-IPL in different regions obtained using three different delineation methods.

[0171] Take the third division method (8 regions) as an example. As shown in Tables 2A and 2B, within the inner ring region with a radius of 1 mm and 3 mm, the average thickness of the nasal, inferior, and temporal regions of the NFL and the volume of the temporal region of the NFL in the elderly group increased significantly. There were no significant changes in other regions of other layers within the inner ring region. Within the outer ring region between 3 mm and 6 mm, the average thickness and volume of the temporal region of the NFL in the elderly group increased significantly. In addition, the average thickness of the nasal, inferior, and temporal regions and the volume of the lower region of the GCL in the elderly group decreased significantly. The average thickness of the upper, nasal, and lower regions of the NFL-IPL and the volume of the upper and nasal regions in the elderly group decreased significantly. The average thickness of the lower region of the GCL-IPL in the elderly group decreased significantly. There were no significant changes in other regions of other layers within the outer ring region.

[0172] In addition, the present invention customized an Area parameter, which represents the ratio of the area with a thickness greater than 45 μm on the GCL thickness map to the area of ​​the 6 mm circle on the ETDRS grid. A two-tailed independent sample t-test revealed a significant difference in Area values ​​between the two age groups (p = 0.032*). A further Pearson bivariate correlation test showed that Area values ​​decreased significantly with age, as shown in Table 2C, confirming the validity of this customized parameter.

[0173] Table 2A Pearson correlation results between structural parameters (inner 3 mm circle) and age

[0174] Table 2B Pearson correlation results between structural parameters (outer 6mm circle) and age

[0175] Table 2 Pearson correlation results between structural parameters (area) and age

[0176] To validate the validity of the defined retinal vascular biomarkers, Pearson bivariate correlation tests were performed between age and vascular parameters. The vascular parameters included general parameters such as the number of nodes, microvascular area abundance, branch point abundance, and fractal abundance. In addition, the corresponding mean, median, variance, skewness, and kurtosis of vascular length, diameter, and tortuosity were also included. As shown in Table 3, all general parameters representing microvascular abundance decreased significantly with age. Other vascular parameters including total vessel length, vessel length abundance, and average vessel length also decreased significantly with age. In addition, the mean and median of vessel diameter and the mean of vessel tortuosity also increased significantly with age.

[0177] Table 3 Pearson correlation results between vascular parameters and age

[0178] To evaluate the validity of the defined retinal function biomarkers, the present invention performed a Pearson bivariate correlation test between age and pERG parameters (including N35, P50, and N95 latency and N35-P50 and P50-N95 amplitudes) for pERG reported results. As shown in Table 4A, with increasing age, P50 latency and P50-N95 amplitudes decreased significantly. Due to the relatively small sample size, the results may be inaccurate.

[0179] However, even with a small number of participants, the validity of functional biomarkers has been demonstrated.

[0180] Table 4A Pearson correlation results between functional parameters (pERG) and age

[0181] In order to evaluate the validity of the defined retinal functional biomarkers, the present invention performs a Pearson bivariate correlation test between age and mfERG parameters (including the latency and amplitude of P1 and N1) on the 103mfERG reported results. The 103mfERG test refers to a test that uses 103 hexagonal stimulation areas in the center of the macular area to obtain electrophysiological responses in different areas of the retina. Specifically, the present invention provides two analysis methods for the reported results of the 103mfERG test, which divide the 103 hexagons into 4 quadrants or 6 concentric rings respectively.

[0182] The first analysis method divided the 103 hexagons into four quadrants, as shown in Figure 13A. Within each quadrant, the software generated an average electrical waveform. Based on this waveform, the amplitude of P1 and N1 and the latency data of P1 and N1 were obtained. Pearson bivariate correlation tests were performed between age and the latency and amplitude of P1 and N1 in each quadrant. As shown in Table 5A, for quadrant 1, the latency of P1 and N1 increased significantly with increasing age; for quadrants 2 and 4, the latency of P1 increased significantly with increasing age; for quadrants 3, there was no significant change in retinal function parameters with increasing age.

[0183] The second analysis method divided the 103 hexagons into six concentric, equidistant rings around the fovea, as shown in Figure 13B. The range of each ring was as follows: Ring 1, 0° to 1°; Ring 2, 1° to 4°; Ring 3, 4° to 8°; Ring 4, 8° to 12°; Ring 5, 12° to 17°; and Ring 6, 17° to 21°. For each ring, the software generated an averaged electrical waveform. Based on this waveform, the amplitudes of P1 and N1, as well as the latencies of P1 and N1, were obtained. Pearson bivariate correlation tests were performed between age and the latencies and amplitudes of P1 and N1 within each ring. As shown in Table 5B, there was no significant change in retinal function with increasing age for Rings 1 and 2; however, there was a significant increase in P1 and N1 latencies with increasing age for Ring 3; a significant increase in P1 latency with increasing age for Rings 4 and 5; and a significant increase in N1 latency with increasing age for Ring 6.

[0184] Table 5A Pearson correlation results between functional parameters (mfERG 4 quadrants) and age

[0185] Table 5B Pearson correlation results between functional parameters (mfERG 6 rings) and age

[0186] Example 7 - Methods for explaining comprehensive analysis

[0187] As demonstrated in Example 6, many retinal imaging and functional characteristics have changed significantly with age, particularly in the outer circular area of ​​the macula. As discussed in Example 5, a comprehensive analysis can be performed by applying two different grid overlays (ETDRS grid and 103 hexagons) to the same macular area centered on the fovea. The results of Example 6 are discussed below. The area within the innermost 1 mm circle of the EDTRS grid corresponds to 0 to 4 degrees on the mfERG report, i.e., the red area. The inner annulus (the area between the 1 mm and 3 mm circles) corresponds to 4 to 8 degrees, i.e., the yellow area. The outer annulus (the area between the 3 mm and 6 mm circles) corresponds to 8 to 12 degrees, i.e., the green area. The analysis results show that within the inner annulus (ring 3), NFL thickness increases significantly with age, and it is in this area that the latencies of N1 and P1 increase significantly with age. Within the outer annulus (ring 4), GCL thickness decreases significantly with age, and in this area, the latency of P1 increases significantly with age. Within a 2-mm-diameter circular area centered on the fovea, vascular abundance significantly decreases with age, and vascular tortuosity also significantly increases with age. No significant age-related retinal changes are observed within the 1-mm circle of the ETDRS grid, as this region lacks vascular structures. Based on these findings, we hypothesize the following biological mechanism underlying retinal changes during aging: With central nervous system aging, retinal vascular abundance decreases dramatically, and vascular tortuosity increases dramatically, indicating rapid vascular hardening. These vascular changes reduce vascular oxygen supply, and retinal ganglion cells, which rely on vascular nutrition, begin to undergo apoptosis due to insufficient nutrition. Ultimately, with fewer ganglion cells transmitting information from the retina to the brain, retinal function deteriorates. These findings suggest a potential mechanism underlying retinal changes during aging in the human central nervous system. Ganglion cells, as projection neurons of the vertebrate retina, are responsible for transmitting information from other retinal neurons to the brain. Their loss can serve as an indicator of neuronal loss within the central nervous system, providing a basis for timely intervention.

[0188] Example 8 - Algorithms and methods for image processing using OpenCV and Shapely in Python

[0189] All patents, patent applications, provisional applications, and publications mentioned or cited herein are incorporated by reference in their entirety, including all figures and tables, to the extent they are not inconsistent with the explicit teachings of this specification.

[0190] It should be understood that the examples and embodiments described herein are for illustrative purposes only, and that those skilled in the art will suggest various modifications or changes thereto, and that such modifications or changes are included within the spirit and scope of the present application. In addition, any element or limitation of any invention or embodiment thereof disclosed herein may be combined with any and / or all other elements or limitations disclosed herein (alone or in any combination) or any other invention or embodiment thereof, and all such combinations are within the scope of the present invention, without limitation thereto.

[0191] References

[0192] Quiroz, JA, & Manji, HK (2002). Enhancing synaptic plasticity and cellular resilience to develop novel, improved treatments for mood disorders. Dialogues in Clinical Neuroscience, 4(1), 73-92. https: / / doi.Org / 10.31887 / dcns.2002.4.1 / jquiroz mood disorders

[0193] Costa,RA,Skaf,M.,Melo Jr,LA,Calucci,D.,Cardillo,JA,Castro,JC,Huang,D.,&Wojtkowski,M.(2006).Retinal assessment using optical coherence tomography.Progress in retinal and eye research,25(3),325-353.https: / / doi.org / 10.1016Zj.preteyeres.2006.03.001

[0194] Le, PH, Patel, BC (2020) Optical Coherence Tomography Angiography. StatPearls Publishing, Treasure Island (FL); PMID: 33085382.

[0195] Fingler, J., Readhead, C., Schwartz, D. M., & Fraser, S. E. (2008). Phase-contrast OCT imaging of transverse flows in the mouse retina and choroid. Investigative ophthalmology & visual science, 49(11), 5055 - 5059. https: / / doi.org / 10.1167 / iovs.07-1627

[0196] Why is Electroretinography Important?LKC Technologies. (2023, January 31). Retrieved March 27, 2023, from https: / / lkc.com / applications / electroretinography-erg-explained /

[0197] Marshak, D. W. (2009). Retinal Ganglion Cells: Anatomy. In Squire, L. R. (Ed.), Encyclopedia of Neuroscience (pp. 211 - 218). Academic Press. https: / / doi.org / 10.1016 / B978-008045046-9.00897-4

[0198] Badwaik, K., Mahmood, K., & Raza, A. (2017). Towards applying OCR and Semantic Web to achieve optimal learning experience. 2017 IEEE 13th International Symposium on Autonomous Decentralized System (ISADS), 262 - 267. https: / / doi.org / 10.1109 / ISADS.2017.40

[0199] Untracht,G.R.,Matos,R.S.,Dikaios,N.,Bapir,M.,Durrani,A.K.,Butsabong,T.,Campagnolo,P.,Sampson,D.D.,Heiss,C.,Sampson,D.M.(2021).OCTAVA:An open-source toolbox for quantitative analysis of optical coherence tomography angiography images.PLoS ONE,16(12),e0261052.https: / / doi.org / 10.1371 / journal.pone.0261052

[0200] Chiu,K.(2Q20).Methods for automated cloud-based quantitative assessment of retinal microvasculature using optical coherence tomography angiography images(H.K.Patent No.63066568).UHK.264P.

[0201] Shiihara,H.,Terasaki,H.,Sonoda,S.,Kakiuchi,N.,Shinohara,Y.,Tomita,M.,&Sakamoto,T.(2017).Objective evaluation of size and shape of superficial foveal avascular zone in normal subjects by optical coherence tomography angiography.Scientific Reports,7(1),9853.https: / / doi.org / 10.1038 / s41598-018-28530-7

[0202] Li,Y.K.,Fung,N.S.-K.,Chan,J.C.H.,Choy,B.N.K.,Chow,L.L.W.,Shih,K.C.,Wong,J.K.W.,&Wong,I.Y.H.(2023).OCTAbiomarkers in adults aged 50 and above:A prospective and cross-sectional community-based study.BMC Ophthalmology,23(1),71.

[0203] https: / / doi.org / 10.1186 / s12886-023-02815-6

[0204] Liu,X.,Lai,S.,Ma,S.,Yang,H.,Liu,L.,Yu,G.,Zhong,S.,Jia,Y.,&Zhong,J.(2022).Development of a Novel Retina-Based Diagnostic Score for Early Detection of Major Depressive Disorder:An Interdisciplinary View.Frontiers in psychiatry,13,897759.https: / / doi.org / 10.3389 / fpsyt.2022.897759

[0205] Hart,N.J.,Koronyo,Y.,Black,K.L.,&Koronyo-Hamaoui,M.(2016).Ocular indicators of Alzheimer's:exploring disease in the retina.Acta neuropathological,132(6), https: / / doi.org / 10.1007 / s00401-016-1613-6

[0206] MacGillivray,T.J.,Trucco,E.,Cameron,J.R.,Dhillon,B.,Houston,J.G.,&van Beek,E.J.(2014).Retinal imaging as a source of biomarkers for diagnosis,characterization and prognosis of chronic illness or long-term conditions.The British journal of radiology,87(1040),20130832.https: / / doi.org / 10.1259 / bjr.20130832

[0207] Hui,J.,Zhao,Y.,Yu,S.,Liu,J.,Chiu,K.,&Wang,Y.(2021).Detection of retinal changes with optical coherence tomography angiography in mild cognitive impairment and Alzheimer’s disease patients:A meta-analysis.PLOS ONE,

[0208] 16(8).https: / / doi.org / 10.1371 / journal.pone.0255362

[0209] Lopez-de-Eguileta, A., Lopez-Garcia, S., Lage, C., Pozueta, A., Garcia-Martinez, M., Kazimierczak, M., Bravo, M., Irure, J., Lopez-Hoyos, M., Munoz-Cac ho,P.,Rodriguez-Perez,N.,Tordesillas-Gutierrez,D.,Goikoetxea,A.,Nebot,C.,Rodriguez-Rodriguez,E.,Casado,A.,10&Sanchez-Juan,P.(2022).The Retinal ganglion cell layer reflects neurodegenerative changes in cognitively unimpaired individuals.Alzheimer's Research&Therapy,14(1),1-13

[0210] Yoon, SP, Grewal, DS, Thompson, AC, Polascik, BW, Dunn, C., Burke, JR, & Fekrat, 15 S. (2019). Retina,3(6), 489-499

[0211] Zhang,M.,Zhong,L.,Han,X.,Xiong,G.,Xu,D.,Zhang,S.,Cheng,H.,Chiu,K.,&Xu,Y.20(2021).Brain and retinal abnormalities in the 5xFAD mouse model of Alzheimer’s disease at early stages.Frontiers in Neuroscience,15,681831.https: / / doi.org / 10.3389 / fnins.2021.681831

[0212] Georgevsky,D.,Retsas,S.,Raoufi,N.,Shimoni,O.,&Golzan,S.M.(2019).A longitudinal assessment of retinal function and structure in the APP / PS1 transgenic mouse model of 25 Alzheimer’s disease.Translational Neurodegeneration,8(30).https: / / doi.org / 10.1186 / s40035-019-0170-z

Claims

1. An integrated analysis method for multimodal retinal ganglion cell (RGC) data for marking central nervous system cell aging, the method comprising: applying an Early Treatment Diabetic Retinopathy Study (ETDRS) grid to a macular plan view reconstructed from an optical coherence tomography (OCT) thickness map to divide the macular plan view into a plurality of regions, wherein the OCT thickness map includes data values ​​for thickness and volume of five different retinal layers, wherein the five different retinal layers include a nerve fiber layer (NFL), a ganglion cell layer (GCL), an inner plexiform layer (IPL), a nerve fiber layer to an inner plexiform layer (NFL-IPL), and a ganglion cell layer to an inner plexiform layer (GCL-IPL); grouping the plurality of regions and determining average thickness and volume data for each region of the five different layers; Optical tomographic angiography (OCTA) was recorded using the posterior pole module and finely sliced ​​to obtain retinal microvascular changes of the OCTA; Obtain a graphic electroretinogram (pERG) report that includes analysis of N35, P50, and N95 latencies and P50 and N95 amplitudes; Obtaining a multifocal electroretinogram (mfERG) report of 103 hexagons of the ETDRS grid with the same center at the macula, the mfERG report includes analysis of the latency of P1 and N1 and the amplitude of P1 and N1; and The presence or absence of RGC aging was identified based on the average thickness and volume data of each area in the five different layers of the OCT thickness map, the retinal microvascular changes of the OCTA, the latencies of N35, P50, and N95 and the amplitudes of P50 and N95 in the pERG report, and the latencies of P1 and N1 and the amplitudes of P1 and N1 in the mfERG report.

2. The method according to claim 1, further comprising: The posterior pole module was used to record the plan view of the macular area reconstructed from the OCT thickness map.

3. The method according to claim 1, wherein: Dividing the macular area plane view into a plurality of regions comprises: dividing the macular area plane view into eight regions, the eight regions comprising an inner ring superior, an inner ring nasal, an inner ring inferior, an inner ring temporal, an outer ring superior, an outer ring nasal, an outer ring inferior, and an outer ring temporal region; and Grouping the plurality of regions comprises: The macular areas are grouped to form 3 annular zones including: the fovea within 1 mm in diameter; an inner annulus between 1 mm and 3 mm in diameter; and an outer annulus between 3 mm and 6 mm in diameter, or, Group the macular areas to form 4 zones including: superior, nasal, inferior, and temporal zones in annular areas between 1 mm and 6 mm, or The macular areas were grouped to form 8 regions including: the inner annular superior, inner annular nasal, inner annular inferior, and inner annular temporal regions with diameters between 1 mm and 3 mm and the outer annular superior, outer annular nasal, outer annular inferior, and outer annular temporal regions with diameters between 3 mm and 6 mm.

4. The method of claim 3, wherein grouping the macular area to form three regions, determining the thickness and volume data of each region of the five different layers comprises: Calculate the average thickness and volume of the inner annular zone area of ​​the five different layers; Calculate the average thickness and volume of the outer annular zone areas of the five different layers; as well as The average thickness and volume of the total annulus, which is the combined area of ​​the inner annulus area and the outer annulus area, is calculated.

5. The method of claim 3, wherein grouping the macular regions to form four regions, determining the thickness and volume data of each region of the five different layers comprises: Calculating the average thickness and volume of the total upper area, the total upper area being the combined area of ​​the inner annular upper area and the outer annular upper area; Calculating the average thickness and volume of the total nose area, the total nose area being the combined area of ​​the inner and outer nose ring areas; Calculating the average thickness and volume of the total lower area, the total lower area being the combined area of ​​the inner lower ring area and the outer lower ring area; as well as The mean thickness and volume of the total temporal area, which is the combined area of ​​the inner and outer annular temporal areas, were calculated.

6. The method of claim 3, wherein grouping the macular regions to form eight regions, determining the thickness and volume data of each region of the five different layers comprises: The average thickness and volume of the inner ring supraregion, inner ring nasal region, inner ring subregion, and inner ring temporal region were calculated in five different layers; as well as The average thickness and volume of the outer annular supraregion, outer annular nasal region, outer annular inferior region, and outer annular temporal region were calculated in five different layers.

7. The method according to claim 1, further comprising: Extracting a region with a thickness greater than 45 μm from the OCT thickness map and calculating the area of ​​the region by an image processing method; Extracting a region within a 6 mm circle from the ETDRS grid and calculating the area of ​​the region by an image processing method; Determining a parameter Area based on the ratio of the area of ​​the region with a thickness greater than 45 μm in the OCT thickness map to the area within a 6 mm circle of the ETDRS grid, wherein the parameter Area is used as a comprehensive indicator of ganglion cell abundance; Whether or not RGC aging exists is determined based on the parameter Area.

8. The method according to claim 1, wherein recording optical tomographic angiography (OCTA) using a posterior pole module and finely stratifying to obtain retinal microvascular changes of the OCTA further comprises: Acquire OCTA vascular image data; The area within the circle with a diameter of 2 mm centered on the macula was extracted as the region of interest (ROI) and this area of ​​the image was cropped; calculating vessel abundance, vessel length, vessel diameter, and vessel tortuosity by applying the OCTAVA method to the ROI; and Calculate the foveal avascular zone (FAZ) parameters, including the FAZ area index and the FAZ circularity index; Identifying whether RGCs are aged based on retinal microvascular changes based on OCTA also includes: The presence or absence of RCG aging was determined based on vessel abundance, vessel length, vessel diameter, and vessel tortuosity as well as foveal avascular zone (FAZ) parameters.

9. The method according to claim 1, wherein: Get a pERG report that also includes: Perform 48-minute pERG test twice and obtain pERG report; and Extracting N35, P50 and N95 latency and P50 and N95 amplitude data from the pERG report by OCR Space text extraction method; The presence or absence of RGC aging was determined based on the latencies of N35, P50, and N95, and the amplitudes of P50 and N95.

10. The method according to claim 1, wherein: Obtaining an mfERG report includes: Perform mfERG test and obtain mfERG report; Divide the 103 hexagons into four different regions / quadrants with one horizontal line and one vertical line, and calculate the average latency and amplitude of P1 and N1 in the four different quadrants, and / or The 103 hexagons were divided into six equally spaced concentric annular regions around the central disk, and the average latency and amplitude of P1 and N1 were calculated for the six different annular regions.

11. A computer program product comprising: A non-transitory computer executable storage device having computer readable program instructions thereon, wherein when the computer readable program instructions are executed by a computer, the computer is caused to perform an integrated analysis method of multimodal retinal ganglion cell (RGC) data for marking central nervous system aging as described in any one of claims 1-10.

12. An integrated analysis device for multimodal retinal ganglion cell (RGC) data for marking central nervous system aging, comprising at least one processor configured to execute the method of any one of claims 1-10.

Images