An artificial intelligence-based mental illness diagnostic system and method using exosome SERS signals.

Abstract

The present invention relates to an AI-based system and method for diagnosing mental illness using exosome SERS signals. The AI-based system for diagnosing mental illness using exosome SERS signals includes: a first learning unit that inputs a first signal map acquired using exosomes acquired from a normal subject and a second signal map acquired using exosomes acquired from a patient with a mental illness into a mental illness diagnostic algorithm and trains the algorithm to classify the exosome SERS signals included in the input signal map as 0 or 1; a signal acquisition unit that drops exosomes acquired from a subject onto a chip containing a plurality of dot arrays and then acquires a signal map including the plurality of exosome SERS signals from the chip; and a diagnosis unit that inputs the acquired signal map into the trained mental illness diagnostic algorithm to acquire a signal value of 0 or 1 for each exosome SERS signal included in the signal map and diagnoses the subject as normal or mentally ill using the average of the acquired signal values.

Description

Technical Field

[0006]

[0001] The present invention relates to a mental disorder diagnosis system and method based on an artificial intelligence platform using exosome SERS signals. More specifically, it detects a SERS signal map across the entire exosome rather than a specific marker, and diagnoses and classifies mental disorders through an artificial intelligence algorithm learned using the detected SERS signal map.

Background Art

[0007] The technology underlying this invention is disclosed in Japanese Patent Publication No. 2021-112167 (published August 5, 2021). [Overview of the project] [Problems that the invention aims to solve]

[0008] Thus, the present invention provides a mental disorder diagnostic system and method that detects a SERS signal map throughout an exosome, which is not a specific marker, and diagnoses and classifies mental disorders through an artificial intelligence algorithm learned using the detected SERS signal map. [Means for solving the problem]

[0009] According to an embodiment of the present invention for solving these technical challenges, an artificial intelligence-based mental illness diagnostic system using exosome SERS signals includes: a first learning unit that inputs a first signal map obtained using exosomes obtained from a healthy person and a second signal map obtained using exosomes obtained from a patient with a mental illness into a mental illness diagnostic algorithm, and trains the mental illness diagnostic algorithm to classify the exosome SERS signals included in the input signal maps as either 0 or 1; a signal acquisition unit that drops exosomes obtained from a subject onto a chip containing multiple dot arrays, and then obtains a signal map containing multiple exosome SERS signals from the chip; and a diagnostic unit that inputs the acquired signal map into a trained mental illness diagnostic algorithm, obtains a signal value of 0 or 1 for each of the exosome SERS signals included in the signal map, and diagnoses the subject as normal or mentally ill using the average of the acquired signal values.

[0010] If a mental illness is diagnosed, the system may further include a classification unit that inputs multiple exosome SERS signals obtained from the subject into multiple mental illness classification algorithms, obtains a signal value of 0 or 1 for each of the multiple exosome SERS signals, and classifies the type of mental illness using the average of the obtained signal values.

[0011] The system may further include a SERS signal acquisition unit that acquires a first signal map from exosomes obtained from healthy individuals and a second signal map from exosomes obtained from patients with mental illness, then labels all n*m (where n and m are identical or distinct natural numbers) exosome SERS signals in the first signal map to 0, and labels all n*m exosome SERS signals in the second signal map to 1.

[0012] The system may further include a second learning unit that inputs a second signal map obtained from patients with mental illnesses that fall under a specific type of mental illness, and a second signal map obtained from patients with mental illnesses excluding those with the specific type of mental illness, into the multiple mental illness classification algorithms, thereby training each mental illness classification algorithm to determine whether or not the input signal map falls under the specific type of mental illness.

[0013] The diagnostic unit inputs the n*m ​​exosome SERS signals included in the acquired signal map into the mental illness diagnostic algorithm, outputs a signal value of 0 or 1 corresponding to each of the n*m ​​exosome SERS signals, and if the average of the output signal values ​​is close to 0, it is classified as normal, and if the average of the output signal values ​​is close to 1, it can be diagnosed as a mental illness.

[0014] The classification unit inputs n*m exosome SERS signals into the plurality of mental disorder classification algorithms, and the plurality of mental disorder classification algorithms output a signal value of 0 or 1 for each of the input n*m exosome SERS signals, and by comparing the average of the output signal values ​​with a classification criterion value for a specific type of mental disorder, it can determine whether or not the disorder falls under that specific type of mental disorder.

[0015] Furthermore, a method for diagnosing a mental illness using a mental illness diagnostic system according to an embodiment of the present invention includes the steps of: inputting a first signal map obtained using exosomes obtained from a healthy person and a second signal map obtained using exosomes obtained from a patient with a mental illness into a mental illness diagnostic algorithm, and training the mental illness diagnostic algorithm to classify the exosome SERS (Surface Enhanced Raman Spectroscopy) signals included in the input signal maps as either 0 or 1; dropping exosomes obtained from a subject onto a chip containing multiple dot arrays, and then obtaining a signal map containing multiple exosome SERS signals from the chip; and inputting the obtained signal map into a trained mental illness diagnostic algorithm to obtain a signal value of 0 or 1 for each of the exosome SERS signals included in the signal map, and diagnosing the subject as normal or mentally ill using the average of the obtained signal values. [Effects of the Invention]

[0016] Thus, according to the present invention, by inputting an exosome SERS signal map into a mental illness diagnostic algorithm, it is possible to diagnose the presence or absence of a mental illness. Furthermore, by inputting the exosome SERS signal map diagnosed as indicating the occurrence of a mental illness into multiple mental illness classification algorithms and re-analyzing it to classify the type of mental illness, a specific diagnosis of the mental illness becomes possible. [Brief explanation of the drawing]

[0017] [Figure 1]It is a configuration diagram for explaining a mental disorder diagnosis system according to an embodiment of the present invention. [Figure 2] It is a flowchart for explaining a mental disorder diagnosis method using the mental disorder diagnosis system according to an embodiment of the present invention. [Figure 3] It is an exemplary diagram for explaining step S210 shown in FIG. 2. [Figure 4] It is an exemplary diagram showing the results of nanoparticle tracking analysis for exosomes of normal individuals and exosomes of patients with depression. [Figure 5] It is an exemplary diagram for explaining a method of labeling an exosome SERS signal in step S220 shown in FIG. 2. [Figure 6] It is an exemplary diagram for explaining a method of learning a mental disorder diagnosis algorithm using the exosome SERS signal labeled in step S220 shown in FIG. 2. [Figure 7] It is an exemplary diagram showing the results of performance evaluation for a mental disorder diagnosis method based on an artificial intelligence platform using an exosome SERS signal according to an embodiment of the present invention.

Mode for Carrying Out the Invention

[0018] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this process, the thickness of the lines and the size of the components shown in the drawings are exaggerated for clarity and convenience of explanation.

[0019] In addition, the terms described below are terms defined in consideration of the functions in the present invention, and these may vary depending on the intention or convention of the user or operator. Therefore, the definitions of such terms must be made based on the content throughout this specification.

[0020] Hereinafter, the mental disorder diagnosis system according to an embodiment of the present invention will be described in more detail with reference to FIG. 1.

[0021] FIG. 1 is a configuration diagram for explaining a mental disorder diagnosis system according to an embodiment of the present invention.

[0022] As shown in FIG. 1, a mental disorder diagnosis system 100 according to an embodiment of the present invention includes a SERS signal collection unit 110, a first learning unit 120, a second learning unit 130, a signal acquisition unit 140, a diagnosis unit 150, and a classification unit 160.

[0023] First, the SERS signal collection unit 110 performs Raman spectroscopy on exosomes collected from the blood serum of normal individuals and exosomes collected from the blood serum of mental disorder patients to obtain respective signal maps.

[0024] Here, the signal map has a size of n*m depending on the number of dot arrays included in the chip. Therefore, the SERS signal collection unit 110 obtains n*m exosome SERS signals depending on the size of the signal map.

[0025] Hereinafter, for convenience of explanation, the signal map obtained from a normal individual is referred to as a first signal map, and the signal map obtained from a mental disorder patient is referred to as a second signal map.

[0026] Then, the SERS signal collection unit 110 labels the n*m exosome SERS signals included in the first signal map as 0, and labels the n*m exosome SERS signals included in the second signal map as 1.

[0027] The first learning unit 120 constructs a mental disorder diagnosis algorithm based on deep learning, and inputs the exosome SERS signals labeled as 0 and the exosome SERS signals labeled as 1 into the constructed mental disorder diagnosis algorithm for learning. Then, the mental disorder diagnosis algorithm outputs a signal value of 0 or 1 corresponding to the input exosome SERS signal.

[0028] The second learning unit 130 constructs a mental illness classification algorithm based on deep learning. In this process, multiple mental illness classification algorithms are formed, corresponding to different types of mental illnesses. Here, the types of mental illnesses may include, but are not necessarily limited to, at least one of depression, bipolar disorder, panic disorder, schizophrenia, dementia, and delusional disorder, and may include a variety of other types of mental illnesses.

[0029] Next, the second learning unit 130 inputs the exosome SERS signals obtained from each patient with a mental illness into multiple mental illness classification algorithms for training. Here, a classification algorithm for a specific type of mental illness is taught to determine whether the exosome SERS signal corresponds to a specific type of mental illness, by inputting the exosome SERS signals obtained from patients with a specific type of mental illness and the exosome SERS signals obtained from patients with the remaining types of mental illnesses excluding the specific type. The mental illness classification algorithm then outputs a signal value of 0 or 1 for the input exosome SERS signal.

[0030] The signal acquisition unit 140 drops exosomes collected from the plasma of the subject onto a chip, and then performs Raman spectroscopy on the chip to acquire multiple exosome SERS signals.

[0031] The diagnostic unit 150 inputs the acquired multiple exosome SERS signals into a pre-trained mental disorder diagnostic algorithm and obtains a signal value of 0 or 1 for each exosome SERS signal. The diagnostic unit 150 then uses the average of the acquired signal values ​​to diagnose the subject as normal or mentally ill.

[0032] If the subject of measurement is diagnosed with a mental disorder, the classification unit 160 inputs multiple exosome SERS signals into multiple pre-trained mental disorder classification algorithms.

[0033] In this way, the multiple mental disorder classification algorithms output signal values ​​of 0 or 1 for the multiple exosome SERS signals input, and the classification unit 160 compares the average of the output signal values ​​with the classification criterion value for the corresponding type of mental disorder to determine whether or not it falls under each type of mental disorder.

[0034] For example, a depression classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a depression classification criterion to determine whether depression is present. A bipolar disorder classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a bipolar disorder classification criterion to determine whether bipolar disorder is present. A panic disorder classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a panic disorder classification criterion to determine whether panic disorder is present. A schizophrenia classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a schizophrenia classification criterion to determine whether schizophrenia is present. The dementia classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a dementia classification criterion to determine whether dementia is present or not. Finally, the delusional disorder classification algorithm outputs a signal value of 0 or 1 for multiple input exosome SERS signals and compares the average of these output signal values ​​to a delusional disorder classification criterion to determine whether delusional disorder is present or not.

[0035] The classification unit 160 then provides information about the mental disorders that have been identified as positive.

[0036] A method for diagnosing mental illness using a mental illness diagnostic system according to an embodiment of the present invention will be described in more detail with reference to Figures 2 to 7.

[0037] Figure 2 is a flowchart illustrating a method for diagnosing mental illness using a mental illness diagnostic system according to an embodiment of the present invention.

[0038] As shown in Figure 2, the method for diagnosing a mental illness using the mental illness diagnosis system according to an embodiment of the present invention includes the steps of training a mental illness diagnosis algorithm and a mental illness classification algorithm, and diagnosing a mental illness using the trained mental illness diagnosis algorithm and mental illness classification algorithm.

[0039] First, to explain the stage of training the mental illness diagnostic algorithm and mental illness classification algorithm, the mental illness diagnostic system 100 collects exosome SERS signals from a group of normal individuals and a group of patients with mental illness (step S210).

[0040] Figure 3 is an illustrative diagram illustrating step S210 shown in Figure 2.

[0041] As shown in Figure 3, plasma samples from healthy individuals are obtained, and exosomes are separated from the obtained plasma using size exclusion chromatography (SEC). Exosomes are also separated from plasma samples from patients with mental illness using the same method.

[0042] Exosomes isolated from plasma samples of patients with mental illness contain both exosomes associated with the mental illness and normal exosomes.

[0043] Figure 4 is an illustrative diagram showing the results of nanoparticle tracking analysis of exosomes from healthy individuals and patients with depression.

[0044] As shown in Figure 4, the results of nanoparticle tracking analysis confirm that there are no significant differences in particle size, mode size, and particle concentration between exosomes from healthy individuals and those from depressed patients. However, heterogeneity was observed in the signals detected by Raman spectroscopy.

[0045] Therefore, according to embodiments of the present invention, exosome SERS signals are detected by performing Raman spectroscopy on exosomes isolated from plasma samples of healthy individuals and exosomes isolated from patients with mental illness.

[0046] To explain this in more detail, exosome solutions obtained by separating them from plasma samples of healthy individuals and exosome solutions obtained by separating them from plasma samples of patients with mental illness are dropped onto their respective Au nanoparticle array chips and then dried.

[0047] Here, the Au nanoparticle aggregate array chip is prepared by precipitating AuNPs (Au nanoparticles) in a colloidal solution, after which the NPs are coated onto an APTES-functionalized glass surface. To enhance the detection rate and uniformity of the signal acquisition process on the APTES-functionalized glass surface, the Au nanoparticle aggregate array chip contains an n*m (where n and m are identical or distinct natural numbers) dot array, and the exosome SERS signal is measured at each dot.

[0048] Next, the SERS signal acquisition unit 110 performs Raman spectroscopy on the dried Au nanoparticle array chip containing the exosome solution to acquire an exosome SERS signal map containing multiple exosome SERS signals corresponding to the dot array.

[0049] Specifically, the SERS signal acquisition unit 110 collects a first signal map containing n*m first exosome SERS signals from the exosome solution of a normal person, and collects a second signal map containing n*m second exosome SERS signals from the exosome solution of a patient with a mental disorder.

[0050] The mental illness diagnostic system 100 according to an embodiment of the present invention constitutes a group of mental illness patients using patients who have been diagnosed with at least one of the following mental illnesses: depression, bipolar disorder, panic disorder, schizophrenia, dementia, and delusional disorder.

[0051] Once step S210 is completed, the first learning unit 120 trains a mental illness diagnostic algorithm using the first signal map obtained from the normal group and the second signal map obtained from the mental illness patient group (step S220).

[0052] Figure 5 is an illustrative diagram illustrating how to label exosome SERS signals in step S220 shown in Figure 2, and Figure 6 is an illustrative diagram illustrating how to learn a mental disorder diagnostic algorithm using the exosome SERS signals labeled in step S220 shown in Figure 2.

[0053] Since the exosome solution from patients with mental illness contains both normal exosomes and mental illness-related exosomes, the n*m ​​dots arranged on the Au nanoparticle array chip may contain only one type of exosome (either normal or mental illness-related), or both. In other words, the multiple exosome SERS signals corresponding to the dot array will each be output differently.

[0054] As shown in Figure 5, the first learning unit 120 according to the embodiment of the present invention does not classify exosome SERS signals based on the presence or absence of mental illness-related exosomes or normal exosomes. Instead, it labels multiple first exosome SERS signals obtained from normal individuals to 0, and multiple second exosome SERS signals obtained from patients with mental illnesses to 1.

[0055] Next, as shown in Figure 6, the first learning unit 120 randomly extracts learning data and test data from the first exosome SERS signal and the second exosome SERS signal.

[0056] The first learning unit 120 then takes the first exosome SERS signal and the second exosome SERS signal corresponding to the extracted learning data as input data, and the labeled values ​​as output data to train a mental illness diagnostic algorithm.

[0057] In other words, the mental illness diagnostic algorithm outputs a signal value of 0 or 1 in response to multiple input exosome SERS signals, and uses the average of these output signal values ​​to first diagnose the presence or absence of a mental illness.

[0058] Next, the second learning unit 130 uses the second exosome SERS signals from the group of patients with mental illness acquired in step S210 to train a mental illness diagnostic algorithm (step S230).

[0059] The second learning unit 130 constructs multiple mental disorder classification algorithms corresponding to depression, bipolar disorder, panic disorder, schizophrenia, dementia, and delusional disorder.

[0060] The second learning unit 130 then inputs the second exosome SERS signals obtained from the depression patient group and the second exosome SERS signals obtained from the remaining groups of mental illness patients (excluding the depression patients) into the depression classification algorithm, and trains the depression classification algorithm to output a signal value of 0 or 1 corresponding to the input exosome SERS signals.

[0061] Furthermore, the second learning unit 130 inputs the second exosome SERS signals obtained from the bipolar disorder patient group and the second exosome SERS signals obtained from the remaining types of mental illness patient groups (excluding bipolar disorder patients) into the panic disorder classification algorithm, and trains the panic disorder classification algorithm to output a signal value of 0 or 1 corresponding to the input exosome SERS signals.

[0062] The second learning unit 130 is used to train the panic disorder classification algorithm, the schizophrenia (mental schizophrenia) classification algorithm, the dementia classification algorithm, and the delusional disorder classification algorithm in the same way.

[0063] Once the algorithm has been trained using step S210 or step S230, the mental illness diagnostic system 100 diagnoses a mental illness in the subject being measured.

[0064] First, the signal acquisition unit 140 acquires the exosome SERS signal extracted from the plasma of the subject being measured (step S240).

[0065] The user collects plasma from the subject and separates exosomes from the collected plasma using chromatography.

[0066] Then, the user drops the solution containing the dissolved exosomes onto an Au nanoparticle array chip and allows it to dry.

[0067] Next, Raman spectroscopy is performed on the Au nanoparticle array chip to obtain n*m (e.g., 100) exosome SERS signals corresponding to the dot array.

[0068] Once step S240 is complete, the diagnostic unit 150 inputs n*m exosome SERS signals into the mental illness diagnostic algorithm to determine whether or not the person being measured has a mental illness (step S250).

[0069] To elaborate, the diagnostic unit 150 inputs n*m exosome SERS signals to the mental illness diagnostic algorithm. The mental illness diagnostic algorithm then outputs a signal value of 0 or 1 corresponding to each of the n*m ​​exosome SERS signals.

[0070] If the average of the output signal values ​​is close to 0, the diagnostic unit 150 diagnoses the subject as a normal person; if the average of the output signal values ​​is close to 1, the diagnostic unit 150 diagnoses the subject as a person with a mental illness.

[0071] In step S250, if the subject is diagnosed with a mental illness, the classification unit 160 inputs n*m exosome SERS signals into multiple pre-learned mental illness classification algorithms to classify the subject's mental illness (step S260).

[0072] In more detail, the classification unit 160 inputs n*m exosome SERS signals into the depression classification algorithm, the manic-depressive disorder classification algorithm, the panic disorder classification algorithm, the schizophrenia (mental schizophrenia) classification algorithm, the dementia classification algorithm, and the delusional disorder classification algorithm, respectively.

[0073] The depression classification algorithm then outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with the depression classification criteria to determine whether or not the subject is diagnosed with depression.

[0074] The bipolar disorder classification algorithm outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with the bipolar disorder classification criteria to determine whether the subject is diagnosed with bipolar disorder.

[0075] The panic disorder classification algorithm outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with the panic disorder classification criteria to determine whether the subject of measurement has panic disorder.

[0076] The schizophrenia classification algorithm outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with the schizophrenia classification criteria to determine whether the subject of measurement falls under the category of schizophrenia.

[0077] The dementia classification algorithm outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with a dementia classification criterion to determine whether the subject of measurement is diagnosed with dementia.

[0078] Finally, the delusional disorder classification algorithm outputs a signal value of 0 or 1 for each of the n*m ​​exosome SERS signals input, and compares the average of these output signal values ​​with the delusional disorder classification criteria to determine whether the subject is diagnosed with delusional disorder.

[0079] Next, the classification unit 160 provides a prediction result for mental illness using the results output from the six mental illness classification algorithms.

[0080] Figure 7 is an illustrative diagram showing the results of a performance evaluation of an artificial intelligence-based method for diagnosing mental disorders using exosome SERS signals according to an embodiment of the present invention.

[0081] As shown in Figure 7a, the human map obtained from exosomes of healthy individuals and the human map obtained from patients with depression are easily distinguishable with the naked eye.

[0082] As shown in Figure 7b, the final diagnostic values ​​were approximately 1.86 times higher in the depression patient group than in the normal group.

[0083] As shown in Figure 7c, the Receiver Operation Characteristic (ROC) curve is shown to verify the effectiveness of the mental illness diagnostic algorithm or mental illness classification algorithm. The area under the curve (AUC) is greater when it is closer to 1, indicating greater usefulness. The mental illness diagnostic algorithm or mental illness classification algorithm according to the embodiment of the present invention has a value of approximately 0.939, with a sensitivity of 91.4% and a specificity of approximately 88.6%. This indicates that the mental illness diagnostic algorithm or mental illness classification algorithm has excellent performance.

[0084] Thus, the mental illness diagnostic system according to the present invention can diagnose the presence or absence of a mental illness by inputting an exosome SERS signal map into a mental illness diagnostic algorithm. Furthermore, by inputting the exosome SERS signal map diagnosed as indicating the presence of a mental illness into multiple mental illness classification algorithms, reanalyzing it, and classifying the type of mental illness, a specific diagnosis of the mental illness becomes possible.

[0085] Although the present invention has been described with reference to the embodiments shown in the drawings, these are merely illustrative, and those skilled in the art will understand that a variety of modifications and equivalent other embodiments are possible. Therefore, the true scope of technical protection of the present invention must be determined by the technical idea of ​​the claims. [Explanation of Symbols]

[0086] 100: Mental Illness Diagnostic System 110: SERS Signal Acquisition Unit 120: First Learning Department 130: Second Learning Department 140: Signal acquisition section 150: Diagnostic Department 160: Classification department

Claims

1. In an artificial intelligence-based mental illness diagnostic system using exosome SERS signals, A first learning unit inputs a first signal map obtained using exosomes acquired from healthy individuals and a second signal map obtained using exosomes acquired from patients with mental illness into a mental illness diagnostic algorithm, and trains the mental illness diagnostic algorithm to classify the exosome SERS signals contained in the input signal maps as either 0 or 1. A signal acquisition unit that, after dropping exosomes obtained from a subject onto a chip containing multiple dot arrays, acquires a signal map containing multiple exosome SERS signals from the chip, and, A diagnostic unit inputs the acquired signal map into a trained mental disorder diagnostic algorithm to obtain a signal value of 0 or 1 for each of the exosome SERS signals included in the signal map, and uses the average of the acquired signal values ​​to diagnose the subject as normal or mentally ill. A mental illness diagnostic system, including...

2. The mental illness diagnostic system according to claim 1, further comprising a classification unit that, when a person is diagnosed with a mental illness, inputs a plurality of exosome SERS signals obtained from the person being measured into a plurality of mental illness classification algorithms, obtains a signal value of 0 or 1 for each of the plurality of exosome SERS signals, and classifies the type of mental illness using the average of the obtained signal values.

3. A mental illness diagnostic system according to claim 2, further comprising a SERS signal acquisition unit that acquires a first signal map from exosomes acquired from healthy individuals and a second signal map from exosomes acquired from patients with mental illness, then labels all n*m (where n and m are the same or different natural numbers) exosome SERS signals included in the first signal map to 0, and labels all n*m exosome SERS signals included in the second signal map to 1.

4. The mental illness diagnostic system according to claim 3, further comprising a second learning unit that inputs a second signal map obtained from mental illness patients who fall under a specific type of mental illness from among the aforementioned mental illness patients, and a second signal map obtained from the remaining mental illness patients excluding those with the specific type of mental illness, into the plurality of mental illness classification algorithms, and trains each mental illness classification algorithm to determine whether or not the signal map input to it falls under the specific type of mental illness.

5. The aforementioned diagnostic unit, The n*m ​​exosome SERS signals included in the acquired signal map are input to the mental illness diagnostic algorithm, and a signal value of 0 or 1 is output corresponding to each of the n*m ​​exosome SERS signals. A mental illness diagnostic system according to claim 3, wherein if the average of the output signal values ​​is close to 0, it is classified as normal, and if the average of the output signal values ​​is close to 1, it is diagnosed as a mental illness.

6. The aforementioned classification unit is A mental illness diagnostic system according to claim 4, comprising: inputting n*m exosome SERS signals into a plurality of mental illness classification algorithms; each of the plurality of mental illness classification algorithms outputting a signal value of 0 or 1 for each of the input n*m exosome SERS signals; and comparing the average of the output signal values ​​with a classification criterion value for a specific type of mental illness to determine whether or not the system corresponds to the specific type of mental illness.

7. A method for diagnosing a mental illness, performed by a mental illness diagnostic system comprising a first learning unit, a signal acquisition unit, and a diagnostic unit, The first learning unit inputs a first signal map obtained using exosomes acquired from healthy individuals and a second signal map obtained using exosomes acquired from patients with mental illness into a mental illness diagnostic algorithm, and trains the mental illness diagnostic algorithm to classify the exosome SERS signals contained in the input signal maps into either 0 or 1. The signal acquisition unit, after dropping exosomes acquired from the subject onto a chip containing multiple dot arrays, acquires a signal map containing multiple exosome SERS signals from the chip, and The diagnostic unit inputs the acquired signal map into a learned mental disorder diagnostic algorithm to acquire a signal value of 0 or 1 for each of the exosome SERS signals included in the signal map, and uses the average of the acquired signal values ​​to diagnose the subject as normal or mentally ill. Methods for diagnosing mental illness, including those mentioned above.

Images