Apparatus and method for prognosis of elderly depression patients

An AI-driven method using brain and blood biomarkers predicts antidepressant treatment effectiveness for elderly depression, addressing the inefficiencies of current treatments by providing personalized and precise care, leading to improved recovery rates and cost reductions.

WO2024232540A1PCT designated stage expired Publication Date: 2024-11-14THE CATHOLIC UNIV OF KOREA IND ACADEMIC COOP FOUND
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
PCT/KR2024/004127
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-08
Filing Date
2024-04-01
Publication Date
2024-11-14

AI Technical Summary

Technical Problem

Current treatments for elderly depression are less effective and more prone to side effects compared to other age groups, leading to prolonged suffering and increased medical and social costs due to the heterogeneous nature of the disease, which requires a more personalized approach considering complex physiological variables.

Method used

An artificial neural network-based apparatus and method that uses brain imaging biomarkers and blood biomarkers to predict the effectiveness of antidepressant treatments, enabling personalized treatment plans by integrating various etiologies of depression through precision medicine.

Benefits of technology

This approach promotes rapid recovery by accurately predicting treatment responses, reducing trial-and-error methods and associated costs, and optimizing treatment outcomes for elderly depression patients.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure KR2024004127_14112024_PF_FP_ABST
    Figure KR2024004127_14112024_PF_FP_ABST
Patent Text Reader

Abstract

Provided according to embodiments of the present invention is an apparatus and method for the prognosis of elderly depression patients, the apparatus comprises: a preprocessing unit for preprocessing brain imaging biomarkers of elderly depression patients; and a prediction unit for predicting a therapeutic effect of an anti-depressive drug in any given elderly depression patient for the brain imaging biomarkers and blood biomarkers or the drug administration effect evaluation results, through the pre-processed brain imaging biomarkers and blood biomarkers or drug administration effect evaluation results together with a prognosis prediction model trained to predict the effectiveness of antidepressant treatment in elderly depression patients by using the pre-processed brain imaging biomarkers and blood biomarkers.
Need to check novelty before this filing date? Find Prior Art

Description

Device and method for predicting the prognosis of elderly patients with depression

[0001] Embodiments of the present disclosure relate to a device and method for predicting the prognosis of elderly patients with depression using an artificial neural network.

[0002] Depression has a wide range of causes, including physiological, psychological, and sociological factors. Depression in the elderly is known to be caused by the interaction of various factors, including amyloid accumulation, neuronal cell destruction and atrophy, cerebrovascular diseases such as stroke, and cerebral inflammatory responses, which are characteristic biological pathologies.

[0003] Despite the highly heterogeneous nature of geriatric depression, current clinical practice applies a homogeneous clinicopathological model to its treatment. This means that all geriatric patients with depression are treated as a single entity, with clinicians prescribing a variety of medications based on their experience until an effective treatment is found.

[0004] However, treatment for geriatric depression is less effective and side effects are more common than for depression in other age groups, requiring more trial and error to find an effective treatment. Until patients with geriatric depression find the right treatment and method, their suffering, associated medical costs, and social and economic losses will continue.

[0005] Depression in the elderly is a disease caused by the complex interaction of various physiological variables, and thus, a disease model shift toward precision medicine that uses integrated information from systems biology is necessary.

[0006] The present embodiments provide a device and method for predicting the prognosis of a patient with depression by predicting the treatment response of geriatric depression using brain imaging and blood biomarkers or drug administration effect evaluation results, thereby promoting rapid recovery of the patient.

[0007] In one aspect, a cytopathology test support device according to one embodiment includes a preprocessing unit for preprocessing a brain imaging biomarker of an elderly patient with depression, and a prediction unit for predicting the antidepressant treatment effect of an elderly patient with depression based on a brain imaging biomarker and a blood biomarker or a drug administration effect evaluation result of any elderly patient with depression through a prognosis prediction model of an elderly patient with depression that is learned to predict the antidepressant treatment effect of an elderly patient with depression using the preprocessed brain imaging biomarker and the blood biomarker together with the preprocessed brain imaging biomarker and the blood biomarker.

[0008] In another aspect, a method for predicting the prognosis of a patient with geriatric depression according to another embodiment includes a preprocessing step of preprocessing a brain imaging biomarker of a patient with geriatric depression, and a prediction step of predicting the effect of antidepressant treatment of a patient with geriatric depression for any brain imaging biomarker and blood biomarker or drug administration effect evaluation result of a prognosis prediction model of a patient with geriatric depression trained to predict the effect of antidepressant treatment of a patient with geriatric depression using the preprocessed brain imaging biomarker and blood biomarker together with the result of evaluating the effect of drug administration effect of the patient with geriatric depression.

[0009] In another aspect, a computer device according to another embodiment includes a memory storing information on preprocessed brain imaging biomarkers and blood biomarkers, information on drug administration effect evaluation results, and a prognosis prediction model for an elderly patient with depression that is learned to predict the effect of antidepressant treatment for an elderly patient with depression using the preprocessed brain imaging biomarkers and blood biomarkers, and a processor that, when a task of predicting the effect of antidepressant treatment for an elderly patient with depression based on the brain imaging biomarkers and blood biomarkers or the drug administration effect evaluation results of any elderly patient with depression is requested, preprocesses the brain imaging biomarkers and executes the prognosis prediction model for an elderly patient with depression stored in the memory to predict the effect of antidepressant treatment for an elderly patient with depression based on the brain imaging biomarkers and blood biomarkers or the drug administration effect evaluation results of any elderly patient with depression.

[0010] The device and method for predicting the prognosis of elderly patients with depression according to the present embodiments can help patients recover quickly by predicting the treatment response of elderly patients with depression using brain imaging and blood biomarkers or drug administration effect evaluation results.

[0011] Figure 1a is a conceptual diagram of a system for predicting the prognosis of an elderly patient with depression according to one embodiment.

[0012] Figure 1b is a graph comparing the prognostic prediction at Tm and the prognostic prediction at Tm+1.

[0013] Figure 2 is a block diagram of an example of a device for predicting the prognosis of an elderly patient with depression in Figure 1.

[0014] Figure 3 illustrates the process of analyzing blood OAB using a ‘multimer detection system (MDS)’.

[0015] Figure 4 is an amyloid-PET image, one of the brain imaging biomarkers.

[0016] Figure 5 illustrates the preprocessing process of T1 MRI, one of the brain imaging biomarkers.

[0017] Figure 6 illustrates the preprocessing process of T2 MRI, one of the brain imaging biomarkers.

[0018] FIG. 7 is a block diagram of another example of a device for predicting the prognosis of an elderly patient with depression of FIG. 1.

[0019] Figure 8 illustrates the process of learning an initial treatment prediction model by preprocessing and analyzing brain MRI T1 and MRI T2 and amyloid-PET images and extracting them as .csv / .txt files.

[0020] Figure 9 illustrates the process of learning a geriatric depression treatment prediction model with blood biomarkers added to the initial treatment prediction model.

[0021] Figure 10 illustrates the process of conducting a blood biomarker test on a total of 80 elderly patients with depression who completed brain imaging tests.

[0022] Figure 11 illustrates T1 and T2 MRI and amyloid-PET analysis using Neurofit's brain imaging automatic diagnosis program 'SegPlus' and 'Scale PET-A'.

[0023] Figure 12 illustrates the process of learning an initial model for predicting treatment of geriatric depression based on brain imaging biomarkers.

[0024] Figure 13 illustrates the process of classifying and collecting blood from elderly patients with depression.

[0025] Figure 14 illustrates the process of enhancing the prediction model by additionally reflecting blood biomarkers in the initial model.

[0026] Figure 15 illustrates the process of finalizing the model (128) after secondary verification through a 6-month longitudinal clinical study following the completion of the first horizontal verification.

[0027] Fig. 16 is a flowchart of a method (200) for predicting the prognosis of an elderly patient with depression according to another embodiment.

[0028] Figure 17 is a configuration diagram of a computing system according to embodiments of the present invention.

[0029] Figure 18 is a configuration diagram of a client-server computer system according to embodiments of the present invention.

[0030] Hereinafter, some embodiments of the present disclosure will be described in detail with reference to exemplary drawings. When adding reference numerals to components in each drawing, identical components may have the same numerals as much as possible even if they are shown in different drawings. In addition, when describing the present embodiments, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the technical idea of ​​the present invention, the detailed description may be omitted. When "includes," "has," "consists of," etc. are used in this specification, other parts may be added unless "only" is used. When a component is expressed in the singular, it may include a plural unless there is a special explicit description.

[0031] Additionally, terms such as first, second, A, B, (a), (b), etc. may be used to describe components of the present disclosure. These terms are only intended to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by the terms.

[0032] In a description of the positional relationship of components, when it is described that two or more components are "connected," "combined," or "connected," it should be understood that the two or more components may be directly "connected," "combined," or "connected," but that the two or more components may also be further "interposed" with another component to be "connected," "combined," or "connected." Here, the other component may be included in one or more of the two or more components that are "connected," "combined," or "connected" to each other.

[0033] In the description of the temporal flow relationship related to components, operation methods, or manufacturing methods, for example, when the temporal or flow relationship is described as “after”, “following”, “next to”, “before”, etc., it may also include cases where it is not continuous, unless “immediately” or “directly” is used.

[0034] Meanwhile, when numerical values ​​or corresponding information (e.g., levels, etc.) for components are mentioned, even without separate explicit description, the numerical values ​​or corresponding information may be interpreted as including an error range that may occur due to various factors (e.g., process factors, internal or external impact, noise, etc.).

[0035] The embodiments are described in detail with reference to the drawings below.

[0036] Figure 1a is a conceptual diagram of a system for predicting the prognosis of an elderly patient with depression according to one embodiment.

[0037] Referring to FIG. 1a, a system (100) for predicting the prognosis of a geriatric patient with depression according to one embodiment integrates various etiologies of geriatric depression using brain imaging biomarkers (20) and blood biomarkers (30) of geriatric patients with depression (10), and then segments the patient based on a prognosis prediction model of the geriatric patient with depression and selects a drug based on the prediction model (130). Since the selected drug is administered to the geriatric patient with depression, personalized treatment is possible.

[0038] A device (120) for predicting the prognosis of a patient with geriatric depression integrates various etiologies of geriatric depression using brain imaging biomarkers (20) and blood biomarkers (30) of geriatric depression patients (10), and then segments the patient based on a prognosis prediction model for geriatric depression patients. Based on the prognosis prediction model for this patient, a drug is selected (130), the drug is administered to the patient, and the results of the drug administration effect evaluation (40) are obtained.

[0039] A device (120) for predicting the prognosis of elderly patients with depression segments patients based on a prognosis prediction model of elderly patients with depression by using brain imaging biomarkers (20) and blood biomarkers (30) of elderly patients with depression (10) as needed, along with the results of previous drug administration effect evaluations (40).

[0040] Specifically, a device (120) for predicting the prognosis of an elderly patient with depression integrates various etiologies of elderly depression using brain imaging biomarkers (20) and blood biomarkers (30) of an elderly patient with depression (10) at the time of the first treatment, for example, at time T1, and then selects a drug (130) based on a prognosis prediction model of an elderly patient with depression, and from the next time point, for example, after time T2, basically selects a drug (130) based on a prognosis prediction model of an elderly patient with depression using the results of previous drug administration effect evaluations (40).

[0041] A device (120) for predicting the prognosis of an elderly patient with depression basically uses the results of the previous drug administration effect evaluation (40) from the next time point, for example, from time point T2, but can additionally use the brain imaging biomarker (20) and blood biomarker (30) of the patient (10) obtained at time point T2.

[0042] Precision medicine integrates the various etiologies of multifactorial diseases and then segments patients (patient stratification) through a new disease model, enabling personalized treatment that predicts which treatment will be most effective for each individual.

[0043] A system (100) for predicting the prognosis of an elderly patient with depression according to one embodiment applies a precision medicine model to an elderly patient with depression (10) to integrate the etiology of the elderly patient with depression (10) through analysis of brain imaging biomarkers (20) and blood biomarkers (10), and then predicts and provides personalized treatment that predicts which antidepressant will be effective for each individual.

[0044] Figure 1b is a graph comparing the prognostic prediction at Tm and the prognostic prediction at Tm+1.

[0045] Referring to Figure 1b, at a specific point in time of depression treatment, for example, Tm (m is a natural number greater than 0), the effect of the treatment progress so far can be measured, the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0046] At the next treatment point, for example, at Tm+1, the treatment progress up to that point can be measured and compared with the previously predicted value, and the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0047] By comparing each prediction at each time point and presenting it to a clinical specialist, optimal treatment can be derived.

[0048] Below, a device (120) for predicting the prognosis of elderly patients with depression is described in detail.

[0049] Figure 2 is a block diagram of a device for predicting the prognosis of an elderly patient with depression in Figure 1.

[0050] Referring to FIGS. 1 and 2, a device (120) for predicting the prognosis of an elderly depression patient includes a preprocessing unit (124) for preprocessing a brain imaging biomarker (20) of an elderly depression patient (10), and a prediction unit (126) for predicting the antidepressant treatment effect of an elderly depression patient for any brain imaging biomarker and blood biomarker or drug administration effect evaluation result of an elderly depression patient through a prognosis prediction model (128) of an elderly depression patient learned to predict the antidepressant treatment effect of an elderly depression patient using the preprocessed brain imaging biomarker (20) and blood biomarker (30) together with the preprocessed brain imaging biomarker (20) and blood biomarker (30) or drug administration effect evaluation result (40).

[0051] Brain imaging biomarkers (20)

[0052] The brain imaging biomarker (20) may be at least one of brain imaging images such as a high-resolution brain gray matter structural MRI (T1 MRI) image, a T2-Weighted Fluid Attenuated Inversion Recovery (T2 FLAIR) image, and a brain amyloid positron emission tomography (amyloid-PET), but is not limited thereto.

[0053] High-resolution brain gray matter structural MRI (hereinafter referred to as T1 MRI) utilizes the magnetization-prepared rapid gradient echo (MPRAGE) imaging method to reconstruct three-dimensional brain images by selecting and extracting cross-sectional images taken at specific intervals, such as 1 mm. This is used to analyze the location and degree of brain atrophy (reduction in thickness and volume) in elderly patients with depression.

[0054] T2 FLAIR (T2-Weighted Fluid Attenuated Inversion Recovery, hereinafter referred to as T2 MRI) is a type of inversion recovery technique that first applies a 180-degree inversion pulse. In order to suppress the signal of cerebrospinal fluid, a T2 FLAIR image with a certain inversion time, for example, an inversion time of approximately 2500 msec, is selected and extracted. The location and degree of cerebrovascular disease (cerebral white matter degeneration) are analyzed through T2 FLAIR image analysis.

[0055] Brain amyloid positron emission tomography (hereinafter referred to as amyloid-PET) is performed by intravenously administering a specific dose of flutemetamol (185 MBq), for example, 10 mCi, to a subject. Images are then acquired using a PET scanner. This allows for the identification of cerebral beta-amyloid pathology. As illustrated in Figure 3, amyloid-PET images can be used to identify the onset of Alzheimer's disease, a cause of dementia.

[0056] Blood biomarkers (30)

[0057] The blood biomarker (30) may be at least one of blood-extracted biomarkers such as blood oligomerized beta-amyloid (hereinafter referred to as OAB), blood neurofilament light chain (hereinafter referred to as NF-L), and blood brain-derived neurotrophic factor (hereinafter referred to as BDNF), but is not limited thereto.

[0058] Oligomerized beta-amyloid (OAB) in blood is analyzed using, for example, the 'Multimer Detection System (MDS)' developed by People Bio Co., Ltd. MDS is an application of sandwich ELISA that confirms antigen-antibody reactions by conjugating enzymes to antibodies. It is possible to selectively detect and analyze only OAB by distinguishing between monomers (one overlapping antigen recognition site) and oligomers (two or more overlapping antigen recognition sites) by configuring the antigen recognition sites of the capture / detection antibodies that react to the antigen to overlap. OAB is known to be an earlier stage lesion than amyloid plaques that can be analyzed through brain amyloid-PET (see Figure 4). In other words, it is an earlier stage lesion than amyloid plaques confirmed in PET images.

[0059] Additionally, in cases where brain amyloid-PET is expensive and brain amyloid-PET imaging is difficult, blood OAB analysis can reflect overall brain amyloid lesions.

[0060] Figure 3 illustrates the process of analyzing blood OAB using a ‘multimer detection system (MDS)’.

[0061] Referring to Figure 3, neurofilament light chain (NF-L) is known to increase in blood and cerebrospinal fluid when brain neurons are damaged or destroyed. It has recently emerged as a biomarker related to the treatment and prognosis of depression, particularly geriatric depression. In this specification, NF-L is analyzed by non-invasively collecting blood.

[0062] Brain-derived neurotrophic factor (BDNF) is a protein in the brain produced by the BDNF gene. It is a member of a group of neurotrophic factors, which are part of the growth factor family. This factor is involved in fundamental nerve growth factors, and BDNF is known to decrease in chronic neuroinflammation. Therefore, it is emerging as a biomarker related to the treatment and prognosis of depression, particularly geriatric depression, by indirectly reflecting neuroinflammation.

[0063] Drug administration effect evaluation results (40)

[0064] The results of the drug administration effectiveness evaluation (40) refer to the results of evaluating the effectiveness of administering medication to elderly patients with depression. For example, the results of the drug administration effectiveness evaluation (40) may be a scorecard that scores the treatment response of elderly patients with depression after administering medication, or an evaluation card that assesses the presence and severity of depression symptoms in elderly patients.

[0065] For example, the drug administration effect evaluation result (40) may be GDS15 (Geriatric depression scale short form 15).

[0066] The Geriatric Depression Scale (GDS) is a tool used to screen for depression in older adults. The Geriatric Depression Scale short form 15 (GDS15) is a shortened version of the GDS, consisting of 15 yes / no questions. These questions are used to assess the presence and severity of depressive symptoms in older adults. The GDS15 is a useful tool for screening for depression in older adults and is also used in research to measure depression outcomes in older adults.

[0067] The preprocessing unit (124) preprocesses the brain imaging biomarker (20) of the elderly depression patient (10). The preprocessing unit (124) can perform preprocessing to segment the region of the image of the brain imaging biomarker (20) of the elderly depression patient (10).

[0068] The prediction model (128) learns and verifies the image images of the brain imaging biomarker (20) preprocessed by dividing the region by the preprocessing unit (124), and for any brain imaging biomarker (20), the image images of the brain imaging biomarker (20) preprocessed by dividing the region by the preprocessing unit (124) and the blood biomarker (30) or the drug administration effect evaluation result (40) can be used to predict the antidepressant treatment effect of an elderly patient with depression. In other words, the prediction model (128) can be trained to predict the antidepressant treatment effect of an elderly patient with depression (10) by using the image images of the brain imaging biomarker (20) preprocessed by dividing the region by the preprocessing unit (124) and the blood biomarker (30) or the drug administration effect evaluation result (40).

[0069] Figure 4 is an amyloid-PET image, one of the brain imaging biomarkers. Figure 5 illustrates the preprocessing process for a T1 MRI, one of the brain imaging biomarkers. Figure 6 illustrates the preprocessing process for a T2 MRI, one of the brain imaging biomarkers.

[0070] Referring to FIGS. 2 to 6, the preprocessing unit (124) preprocesses the T1 MRI image. Specifically, the preprocessing unit (124) removes the skull from the T1 MRI image along with "pixel size-actual length normalization." In the coronal section of the frontal lobe, the cut that appears most straight from the basilar artery is detected. Only the temporal lobe is recognized and cropped, and the space is classified into, for example, the hippocampus (①), STL (②), MTL (③), ITL (④), and insula (⑤).

[0071] The preprocessing unit (124) preprocesses the T2 MRI image. Specifically, the preprocessing unit (124) removes the skull from the T2 MRI image along with “pixel size-actual length standardization.” In the transverse section, two cuts are detected: a cut (1) in which the ventricle is captured the largest and a cut (2) facing the upper end. The preprocessing unit (124) classifies the space in the first image into, for example, the ant.periventricular cut (①), middle.periventricular cut STL (②), and post.periventricular cut (③) of the right periventricular area and the ant.periventricular cut (④), middle.periventricular cut (⑤), and post.periventricular cut (⑥) of the left periventricular area, and uses the second image as is.

[0072] Referring back to FIG. 2, the prediction model (128) can be trained using preprocessed brain imaging biomarkers (20) and blood biomarkers (30), or can be trained using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) together with drug administration effect evaluation results (40).

[0073] As described above, the device (120) for predicting the prognosis of a patient with geriatric depression can select a drug based on a prognosis prediction model of a patient with geriatric depression after integrating various etiologies of geriatric depression using brain imaging biomarkers (20) and blood biomarkers (30) of the patient with geriatric depression (10) at the time of the first treatment, for example, at time T1.

[0074] When the prediction model (128) is trained using the preprocessed brain imaging biomarker (20) and blood biomarker (30), the prediction model (128) can be trained as an initial model for predicting treatment of geriatric depression based on the brain imaging biomarker (20), and can be trained as a model for predicting treatment of geriatric depression using the blood biomarker (30) together with the brain imaging biomarker (20).

[0075] From the next point in time, for example, after the T2 point in time, the drug can be selected based on the prognosis prediction model for elderly patients with depression by basically using the results of the previous drug administration effect evaluation (40).

[0076] A device (120) for predicting the prognosis of an elderly patient with depression basically uses the results of the previous drug administration effect evaluation (40) from the next time point, for example, from time point T2, but can additionally use the brain imaging biomarker (20) and blood biomarker (30) of the patient (10) obtained at time point T2.

[0077] As illustrated in FIG. 2, when the prediction model (128) is trained using the preprocessed brain imaging biomarkers (20) and blood biomarkers (30) together with the drug administration effect evaluation results (40), if only the drug administration effect evaluation results of any elderly depression patient are provided, the prediction unit (126) can predict the antidepressant treatment effect of the elderly depression patient for the previously used brain imaging biomarkers (20) and blood biomarkers (30) together with the drug administration effect evaluation results (40) of any elderly depression patient through the prognosis prediction model (128) of the elderly depression patient.

[0078] When a new brain imaging biomarker (20) and a blood biomarker (30) are provided together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient, the prediction unit (126) can predict the antidepressant treatment effect of an elderly depression patient for the new brain imaging biomarker (20) and blood biomarker (30) together with the results of the drug administration effect evaluation of an arbitrary elderly depression patient through a prognosis prediction model of the elderly depression patient.

[0079] FIG. 7 is a block diagram of another example of a device for predicting the prognosis of an elderly patient with depression of FIG. 1.

[0080] Referring to FIG. 7, the prediction model (128) may include a first prediction model (128a) learned using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) and a second prediction model (128b) learned using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) along with drug administration effect evaluation results (40).

[0081] If only the brain imaging biomarker (20) and blood biomarker (30) of an elderly patient with depression are provided, the prediction unit (126) can predict the antidepressant treatment effect of the elderly patient with depression through the first prediction model (128a).

[0082] When only the drug administration effect evaluation results (40) are provided or when new brain imaging biomarkers (20) and blood biomarkers (30) are provided together with the drug administration effect evaluation results (40), the prediction unit (126) can predict the antidepressant treatment effect of elderly patients with depression through the second prediction model (128b).

[0083] For example, if only the results of the drug administration effect evaluation (40) for an arbitrary elderly patient with depression are provided, the prediction unit (126) can predict the antidepressant treatment effect for the elderly patient with depression for the previously used brain imaging biomarker (20) and blood biomarker (30) together with the results of the drug administration effect evaluation for the arbitrary elderly patient with depression through the second prediction model (128b).

[0084] At this time, the previously used brain imaging biomarker (20) and blood biomarker (30) may be brain imaging biomarkers (20) and blood biomarkers (30) obtained at the time of the previous treatment, but are not limited thereto and may be brain imaging biomarkers (20) and blood biomarkers (30) obtained at the time of the previous treatment.

[0085] When a new brain imaging biomarker (20) and a blood biomarker (30) are provided together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient, the prediction unit (126) can predict the antidepressant treatment effect of the elderly depression patient for the new brain imaging biomarker (20) and blood biomarker (30) together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient through the second prediction model (128b).

[0086] As described above with reference to Figure 1b, at a specific point in time of depression treatment, for example, Tm (where m is a natural number greater than 0), the effect of the treatment progress so far can be measured, and the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0087] At the next treatment point, for example, at Tm+1, the treatment progress up to that point can be measured and compared with the previously predicted value, and the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0088] By comparing each prediction at each time point and presenting it to a clinical specialist, optimal treatment can be derived.

[0089] Development of an initial model for predicting treatment of geriatric depression based on brain imaging biomarkers.

[0090] Extracted brain MRI T1 and MRI T2 and amyloid-PET images are acquired, refined, processed and preprocessed to analyze the degree and area of ​​brain atrophy, the degree and area of ​​cerebral white matter degeneration, and the degree and area of ​​brain amyloid deposition, respectively. The analyzed results can be extracted as .csv or .txt files and immediately used in a prediction model, which is used to train an initial model (128a) for predicting treatment of geriatric depression (see Fig. 8).

[0091] Figure 8 illustrates the process of learning the initial treatment prediction model (128a) by preprocessing and analyzing brain MRI T1 and MRI T2 and amyloid-PET images and extracting them as .csv / .txt files.

[0092] Development of a treatment prediction model for geriatric depression using blood biomarkers along with brain imaging biomarkers.

[0093] By modifying and supplementing the developed initial model (128a), a model (128) for predicting treatment of geriatric depression that can reflect both brain imaging biomarkers (20) and blood biomarkers (30) is learned (see Fig. 9).

[0094] Figure 9 illustrates the process of learning a geriatric depression treatment prediction model (128) with a blood biomarker (30) added to the initial treatment prediction model (128a).

[0095] As illustrated in Fig. 9, the geriatric depression treatment prediction model (128) can be trained using a deep learning model.

[0096] In this specification, a deep learning model may be a model that stacks artificial neural networks in multiple layers. A deep learning model can be implemented as a model that automatically learns the features of each image by training a deep neural network composed of multiple layers of data, and then trains the network to minimize the error in the objective function, i.e., prediction accuracy.

[0097] For example, a deep learning model might be a convolutional neural network (CNN). CNNs can be implemented in various schemes, such as the Visual Geometry Group (VGG) network, Inception (GoogleNet), ResNet, and DenseNet.

[0098] However, the deep learning model described in the embodiments of the present invention is not limited to CNN, and other types of deep learning models (e.g., DHN (Deep Hierarchical Network), CDBN (Convolutional Deep Belief Network), DDN (Deconvolutional Deep Network), RNN (Recurrent Neural Network)) that can be used now or in the future may also be used.

[0099] Deep learning models can be implemented using deep learning frameworks. Deep learning frameworks provide commonly used functions in the development of deep learning models in the form of libraries and support the efficient use of system software and hardware platforms. In this embodiment, the deep learning model can be implemented using any deep learning framework currently available or to be released in the future.

[0100] Training a deep learning model involves adjusting its parameters so that its predicted values ​​for input data are as similar as possible to the ground truth. This process of repeatedly inputting training data into a deep learning model, comparing the model's predicted values ​​with the ground truth, and adjusting the model's parameters can improve the model's prediction accuracy.

[0101] Meanwhile, the learning process of a deep learning model can be expressed as training, and its result can be expressed as learning, but either training or learning can be used to express the learning process or its result.

[0102] During the training process for a deep learning model, the parameters of the deep learning model can be adjusted in a way that minimizes the loss function value for the input.

[0103] A loss function is a function that defines the error between the predicted value and the actual value. Examples of loss functions include the mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE), binary cross-entropy, or categorical cross-entropy.

[0104] Meanwhile, the process of repeatedly training to improve the predictive accuracy of deep learning models can lead to the problem of overconfidence. Deep learning models optimized to improve predictive accuracy for inputs used in training can actually lead to inaccurate predictions for actual inputs. To account for the uncertainty inherent in deep learning models, work is needed to prevent overconfidence during the training process.

[0105] Advancement of a Predictive Model for Geriatric Depression Treatment through Clinical Validation

[0106] The developed prediction model (128) is subjected to a first supplementary task by conducting a horizontal verification on geriatric depression patients (10) who have completed treatment. Thereafter, the prediction model (128) is applied to geriatric depression patients (10) before treatment is provided, and a 6-month follow-up observation is conducted to confirm the actual treatment response, thereby ultimately improving the prognosis prediction model (128).

[0107] The results predicted before the start of treatment by the prediction model (128) can be classified into, for example, antidepressant monotherapy, atypical antipsychotic monotherapy, antidepressant + antidepressant treatment, antidepressant + atypical antipsychotic treatment, and antidepressant + basic controller treatment, but are not limited thereto.

[0108] In other words, the prediction unit (126) can classify the antidepressant treatment of an elderly patient with depression (10) into one of antidepressant monotherapy, atypical antipsychotic monotherapy, antidepressant + antidepressant treatment, antidepressant + atypical antipsychotic drug treatment, and antidepressant + basic regulator treatment through the learned prognosis prediction model (128) of an elderly patient with depression, but is not limited thereto.

[0109] Experimental example

[0110] Among the three blood biomarkers, blood OAB analysis was performed using People Bio Co., Ltd.'s 'Multimer Detection System (MDS)', and blood BDNF and NF-L were performed through contract testing in conjunction with People Bio Co., Ltd.

[0111] Recruitment of subjects for the development of a treatment prediction model for geriatric depression (actual study period: 6 months)

[0112] A total of 80 elderly patients with depression who had completed brain imaging tests were selected from a database of over 5,000 elderly patients maintained by the institution (treatment-responsive group (N=40); refractory group (N=40)) who responded well to treatment. Blood biomarkers (amyloid-beta oligomer (OAB), neurofilament (NF-L), and brain-derived neurotrophic factor (BDNF)) were tested in the selected patients (see Fig. 10).

[0113] Figure 10 illustrates the process of conducting a blood biomarker test on a total of 80 elderly patients with depression who completed brain imaging tests.

[0114] For brain MRI, the brain atrophy and brain white matter automatic analysis program 'SegPlus' jointly developed by our research team and Neurofit Inc. was used, and for brain amyloid-PET, the PET automatic analysis program 'Scale PET-A' was used to analyze the area and degree of brain atrophy, the area and degree of brain white matter abnormality, and the degree of amyloid deposition in each brain region, respectively (see Figure 11).

[0115] Figure 11 illustrates T1 and T2 MRI and amyloid-PET analysis using Neurofit's brain imaging automatic diagnosis program 'SegPlus' and 'Scale PET-A'.

[0116] Development of an initial model for predicting treatment of geriatric depression using brain imaging and recruitment of additional subjects.

[0117] An initial model (brain imaging-based initial prediction model) capable of predicting drug response was developed using brain MRI and amyloid-PET of the selected treatment response group (N=40) and the refractory patient group (N=40) (see Fig. 12).

[0118] Figure 12 illustrates the process of learning an initial model for predicting treatment of geriatric depression based on brain imaging biomarkers.

[0119] We recruited an additional 200 elderly patients with depression from the Catholic Neuroimaging Database. The 200 patients were divided into three groups: those who responded to selective serotonin reuptake inhibitor (SSRI) medications (SSRI group), those who responded to medications other than SSRIs (non-SSRI group), those taking two or more antidepressants (combination group), and those who were difficult to treat and required antidepressants and mood stabilizers / atypical antipsychotics (refractory group). After recruitment, blood biomarker tests (OAB, NF-L, BDNF) were performed (see right side of Figure 13).

[0120] Figure 13 illustrates the process of classifying and collecting blood from elderly patients with depression.

[0121] Development of a predictive model for geriatric depression treatment using brain imaging and blood biomarkers through advancement of the initial model.

[0122] The developed brain imaging-based initial prediction model (128a) was refined by reflecting the blood biomarkers (30) of 280 people. Through the refinement, a prediction model (128) for geriatric depression treatment that can reflect both brain imaging biomarkers (20) and blood biomarkers (30) was developed. Through the developed prediction model (128), a model was developed that can predict geriatric depression patients as needing SSRIs, needing drugs other than SSRIs, needing two or more antidepressants (needing multiple antidepressants), and intractable patients needing antidepressants and atypical antipsychotics / mood stabilizers (see Fig. 14).

[0123] Figure 14 illustrates the process of enhancing the prediction model (128) by additionally reflecting a blood biomarker (30) in the initial model (128a).

[0124] Fifty elderly patients with depression who had completed brain imaging were selected from the Catholic Neuroimaging Database. Blood biomarker testing was performed on these selected patients. The lateral accuracy of the developed neuroimaging biomarker (20) and blood biomarker (30)-based geriatric depression treatment prediction model (128) was verified, and the model was modified based on this data.

[0125] Validation and final refinement through longitudinal clinical validation of a prediction model for geriatric depression treatment.

[0126] The developed geriatric depression treatment prediction model (128) utilizing brain imaging biomarkers (20) and blood biomarkers (30) requires further validation in actual clinical patients and further supplementation. For clinical validation, 50 geriatric depression patients were selected and followed up for 6 months. After 6 months of longitudinal observation, the longitudinal accuracy of the geriatric depression treatment prediction model (128) was verified, and errors were corrected based on this, ultimately completing the prognosis prediction model (128) (see Figure 15).

[0127] Figure 15 illustrates the process of finalizing the model (128) after secondary verification through a 6-month longitudinal clinical study following the completion of the first horizontal verification.

[0128] Finally, the geriatric depression treatment prediction model (128) using the brain imaging biomarkers (20) and blood biomarkers (30) developed together with the results of the drug administration effect evaluation (40) was verified in actual clinical patients and then supplemented. For example, 50 geriatric depression patients were selected for clinical verification and followed up for 6 months. After the 6-month longitudinal observation, the longitudinal accuracy of the geriatric depression treatment prediction model (128) was verified, and errors were corrected through this to finally complete the prognosis prediction model (128).

[0129] With reference to the above drawings, a device (120) for predicting the prognosis of a geriatric patient with depression and a system (100) including the same have been described. Hereinafter, with reference to the drawings, a method (200) for predicting the prognosis of a geriatric patient with depression according to another embodiment will be described in detail.

[0130] Fig. 16 is a flowchart of a method (200) for predicting the prognosis of an elderly patient with depression according to another embodiment.

[0131] Referring to FIG. 16, a method (200) for predicting the prognosis of an elderly patient with depression according to another embodiment includes a preprocessing step (S210) of preprocessing a brain imaging biomarker (20) of an elderly patient with depression (10), a prediction step (S220) of predicting the effect of antidepressant treatment of an elderly patient with depression for any brain imaging biomarker and blood biomarker or drug administration effect evaluation result of an elderly patient with depression using a prognosis prediction model (128) of an elderly patient with depression that is trained to predict the effect of antidepressant treatment of an elderly patient with depression using the preprocessed brain imaging biomarker (20) and blood biomarker (30) together with the preprocessed brain imaging biomarker (20) and blood biomarker (30) or drug administration effect evaluation result (40).

[0132] As described above, the brain imaging biomarker may be, but is not limited to, at least one of a high-resolution brain gray matter structural MRI (T1 MRI) image, a T2-Weighted Fluid Attenuated Inversion Recovery (T2 FLAIR) image, and a brain amyloid positron emission tomography (amyloid-PET).

[0133] As described above, the blood biomarker may be at least one of, but is not limited to, blood oligomerized beta-amyloid (OAB), blood neurofilament light chain (NF-L).

[0134] In the preprocessing step (S210), the image images of brain imaging biomarkers are segmented into specific regions and preprocessed. The prediction model (128) can be trained to predict the treatment effect of antidepressants in elderly patients with depression using the image images of brain imaging biomarkers, which are segmented into specific regions and preprocessed in the preprocessing step (S210), and the results of blood biomarker or drug administration effect evaluation (40).

[0135] As described above with reference to FIG. 2, the prediction model (128) can be trained using preprocessed brain imaging biomarkers (20) and blood biomarkers (30), or can be trained using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) together with drug administration effect evaluation results (40).

[0136] As described above, the method (200) for predicting the prognosis of a patient with geriatric depression can select a drug based on a prognosis prediction model of a patient with geriatric depression after integrating various etiologies of geriatric depression using brain imaging biomarkers (20) and blood biomarkers (30) of the patient with geriatric depression (10) at the time of the first treatment, for example, at time T1.

[0137] As described above, when the prediction model (128) is trained using the preprocessed brain imaging biomarker (20) and blood biomarker (30), the prediction model (128) is trained as an initial model (128a) for predicting geriatric depression treatment based on the brain imaging biomarker, and is trained as a model (128) for predicting geriatric depression treatment using the blood biomarker (30) together with the brain imaging biomarker (20).

[0138] From the next time point, for example, after time point T2, the drug can be selected based on the prognosis prediction model for elderly patients with depression by basically using the results of the previous drug administration effect evaluation (40).

[0139] A method (200) for predicting the prognosis of an elderly patient with depression basically uses the results of the previous drug administration effect evaluation (40) from the next time point, for example, from time point T2, but can additionally use the brain imaging biomarker (20) and blood biomarker (30) of the patient (10) obtained at time point T2.

[0140] As illustrated in FIG. 2, when the prediction model (128) is trained using the preprocessed brain imaging biomarkers (20) and blood biomarkers (30) together with the drug administration effect evaluation results (40), if only the drug administration effect evaluation results of any elderly depression patient are provided, in the prediction step (S220), the antidepressant treatment effect of the elderly depression patient can be predicted for the previously used brain imaging biomarkers (20) and blood biomarkers (30) together with the drug administration effect evaluation results (40) of any elderly depression patient through the prognosis prediction model (128).

[0141] When a new brain imaging biomarker (20) and a blood biomarker (30) are provided together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient, in the prediction step (S220), the antidepressant treatment effect of an elderly depression patient can be predicted for the new brain imaging biomarker (20) and blood biomarker (30) together with the results of the drug administration effect evaluation of an arbitrary elderly depression patient through a prognosis prediction model of the elderly depression patient.

[0142] As described above, the prediction model (128) may include a first prediction model (128a) learned using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) and a second prediction model (128b) learned using preprocessed brain imaging biomarkers (20) and blood biomarkers (30) along with drug administration effect evaluation results (40).

[0143] If only the brain imaging biomarker (20) and blood biomarker (30) of an elderly patient with depression are provided, the antidepressant treatment effect of the elderly patient with depression can be predicted through the first prediction model (128a) in the prediction step (S220).

[0144] When only the drug administration effect evaluation results (40) are provided or new brain imaging biomarkers (20) and blood biomarkers (30) are provided together with the drug administration effect evaluation results (40), the antidepressant treatment effect of elderly patients with depression can be predicted through the second prediction model (128b) in the prediction step (S220).

[0145] For example, if only the results of the drug administration effect evaluation (40) for an arbitrary elderly depression patient are provided, in the prediction step (S220), the second prediction model (128b) can be used to predict the antidepressant treatment effect for an elderly depression patient based on previously used brain imaging biomarkers (20) and blood biomarkers (30) together with the results of the drug administration effect evaluation for an arbitrary elderly depression patient.

[0146] At this time, the previously used brain imaging biomarker (20) and blood biomarker (30) may be brain imaging biomarkers (20) and blood biomarkers (30) obtained at the time of the previous treatment, but are not limited thereto and may be brain imaging biomarkers (20) and blood biomarkers (30) obtained at the time of the previous treatment.

[0147] When a new brain imaging biomarker (20) and a blood biomarker (30) are provided together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient, in the prediction step (S220), the second prediction model (128b) can be used to predict the antidepressant treatment effect of an elderly depression patient for the new brain imaging biomarker (20) and blood biomarker (30) together with the results of the drug administration effect evaluation (40) of an arbitrary elderly depression patient.

[0148] As described above with reference to Figure 1b, at a specific point in time of depression treatment, for example, Tm (where m is a natural number greater than 0), the effect of the treatment progress so far can be measured, and the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0149] At the next treatment point, for example, at Tm+1, the treatment progress up to that point can be measured and compared with the previously predicted value, and the long-term effect can be predicted when the existing treatment is maintained, and the long-term effect can be predicted when multiple new treatments are introduced.

[0150] By comparing each prediction at each time point and presenting it to a clinical specialist, optimal treatment can be derived.

[0151] In the prediction step (S220), the learned prognosis prediction model (128) of elderly depression patients can be used to classify the treatment of elderly depression patients (10) into one of the following: antidepressant monotherapy, atypical antipsychotic monotherapy, antidepressant + antidepressant treatment, antidepressant + atypical antipsychotic treatment, or antidepressant + basic regulator treatment.

[0152] The method (200) for predicting the prognosis of an elderly patient with depression according to another embodiment described with reference to FIG. 15 can apply the same or similar contents as those described in the device (120) for predicting the prognosis of an elderly patient with depression according to one embodiment described with reference to FIGS. 1 to 14.

[0153] The device (120) and method (200) for predicting the prognosis of an elderly patient with depression described with reference to the above drawing have the following effects.

[0154] ① By developing AI-based automatic analysis technologies for image indicators related to geriatric depression, we can assist specialists in image interpretation and diagnosis, and expand the application cases of such technologies.

[0155] ② Rather than simply providing assistance with image interpretation, it is possible to promote rapid recovery of patients by predicting the treatment response of geriatric depression using brain imaging and blood biomarkers or drug administration effect evaluation results.

[0156] ③ Reducing socioeconomic costs associated with geriatric depression: Medical costs for geriatric depression are steadily increasing. Assuming treatment could prevent 10% of the worsening of geriatric depression, combined with medical and social costs, this could result in a national savings of KRW 400 billion in annual geriatric depression treatment costs. Furthermore, geriatric depression is a risk factor for dementia, and assuming treatment could prevent 10% of dementia worsening, the national dementia treatment cost could be reduced by an additional KRW 1 trillion annually.

[0157] Although the device (120) and method (200) for predicting the prognosis of a geriatric depression patient according to embodiments have been described with reference to the above drawings, the present invention is not limited thereto. For example, it has been described that the prognosis of a geriatric depression patient is predicted using brain imaging biomarkers, blood biomarkers, and drug administration effect evaluation results of the geriatric depression patient. However, if at least one of the brain imaging biomarkers, blood biomarkers, and drug administration effect evaluation results of the geriatric depression patient is not provided, information about them may be used as fixed or virtual values.

[0158] As another example, the present invention describes training a predictive model using both brain imaging biomarkers and blood biomarkers of elderly patients with depression, and predicting the prognosis of elderly patients with depression based on both brain imaging biomarkers and blood biomarkers in an arbitrary patient. However, this is not limited to this. For example, a predictive model can be trained using only one of the brain imaging biomarkers and blood biomarkers of elderly patients with depression. In this case, the prognosis of elderly patients with depression can also be predicted based on either the brain imaging biomarker or blood biomarker of an arbitrary patient.

[0159] Figure 17 is a configuration diagram of a computing system (300) according to embodiments of the present invention.

[0160] Referring to FIG. 17, a computing system (300) may include a memory (310) and a processor (320).

[0161] The memory (310) may store information on preprocessed brain imaging biomarkers (20) and blood biomarkers (30), information on drug administration effect evaluation results (40), and a prognosis prediction model (128) for elderly depression patients that is trained to predict the antidepressant treatment effect in elderly depression patients using the preprocessed brain imaging biomarkers (20) and blood biomarkers (30) or drug administration effect evaluation results (40). The memory (310) may be a volatile memory (e.g., SRAM, DRAM) or a nonvolatile memory (e.g., NAND Flash).

[0162] When a task of predicting the effect of antidepressant treatment on an elderly patient with depression (10) based on the brain imaging biomarker (20) and blood biomarker (30) or the drug administration effect evaluation result of an arbitrary elderly patient with depression (10) is requested, the processor (320) preprocesses the brain imaging biomarker (20) and executes the learned prognosis prediction model (128) of an elderly patient with depression stored in the memory (310) to predict the effect of antidepressant treatment on an elderly patient with depression based on the brain imaging biomarker (20) and blood biomarker (30) or the drug administration effect evaluation result (40) of an arbitrary elderly patient with depression (10).

[0163] A computing system according to embodiments of the present invention may include a computer device (300) including a memory (310) and a processor (320) and a server (400) including a memory (410) and a processor (420). The computer device (300) and the server (400) may be connected wiredly or wirelessly through a network.

[0164] The memory (410) of the server (400) can store a prognosis prediction model (128) of an elderly depression patient that has been trained to predict the antidepressant treatment effect of an elderly depression patient (10) using the preprocessed brain imaging biomarker (20) and blood biomarker (30) or drug administration effect evaluation result (40) described above.

[0165] When a task of predicting the effect of antidepressant treatment on an elderly patient with depression based on the brain imaging biomarker (20) and blood biomarker (30) or the drug administration effect evaluation result (40) of an arbitrary elderly patient with depression is requested (query), the processor (320) of the computer device (300) preprocesses the brain imaging biomarker (20). The memory (310) of the computer device (300) can store the brain imaging biomarker (20) and blood biomarker (30) or the drug administration effect evaluation result (40) of the aforementioned arbitrary elderly patient with depression (10).

[0166] The processor (320) of the computer device (300) can transmit the brain imaging biomarker (20) and blood biomarker (30) or drug administration effect evaluation result (40) of any elderly depression patient (10) stored in the memory (310) and this request (query) to the server (400).

[0167] The processor (420) of the server (400) can execute a learned prognosis prediction model for elderly depression patients to predict the antidepressant treatment effect of an elderly depression patient based on the brain imaging biomarker (20) and blood biomarker (30) or drug administration effect evaluation result (40) of any elderly depression patient, and transmit the result to a computer device (300).

[0168] Various possible examples are described below for the computer system described with reference to FIGS. 17 and 18.

[0169] The device (120) for predicting the prognosis of a patient with geriatric depression may be configured as a computing system (300) as shown in FIG. 17, or may be configured as a GPU server equipped with a GPU processor and general memory and storage for storing scan files (WSI images), but the present invention is not limited thereto.

[0170] The device (120) for predicting the prognosis of a patient with geriatric depression described above may be implemented by a computing device including at least some of a processor, a memory, a user input device, and a presentation device. The memory is a medium that stores computer-readable software, applications, program modules, routines, instructions, and / or data, etc., which are coded to perform a specific task when executed by the processor. The processor can read and execute the computer-readable software, applications, program modules, routines, instructions, and / or data stored in the memory. The user input device may be a means for allowing a user to input a command to cause the processor to perform a specific task or to input data necessary for the execution of a specific task. The user input device may include a physical or virtual keyboard or keypad, key buttons, a mouse, a joystick, a trackball, a touch-sensitive input device, or a microphone. The presentation device may include a display, a printer, a speaker, or a vibration device.

[0171] Computing devices can include a variety of devices, including smartphones, tablets, laptops, desktops, servers, and clients. A computing device may be a single, standalone device, or it may include multiple computing devices operating in a distributed environment, each of which collaborates with another through a communications network.

[0172] In addition, the device (120) for predicting the prognosis of a geriatric depression patient described above can be executed by a computing device having a processor and a memory storing computer-readable software, applications, program modules, routines, instructions, and / or data structures coded to perform a method (200) for predicting the prognosis of a geriatric depression patient using a deep learning model when executed by the processor.

[0173] The embodiments described above may be implemented through various means. For example, the embodiments may be implemented through hardware, firmware, software, or a combination thereof.

[0174] In the case of hardware implementation, the method (200) for predicting the prognosis of elderly depression patients using a deep learning model according to the present embodiments can be implemented by one or more ASICs (Application Specific Integrated Circuits), DSPs (Digital Signal Processors), DSPDs (Digital Signal Processing Devices), PLDs (Programmable Logic Devices), FPGAs (Field Programmable Gate Arrays), processors, controllers, microcontrollers, or microprocessors.

[0175] For example, a method (200) for predicting the prognosis of a patient with geriatric depression according to embodiments can be implemented using an artificial intelligence semiconductor device in which neurons and synapses of a deep neural network are implemented using semiconductor devices. In this case, the semiconductor devices may be currently used semiconductor devices, such as SRAM, DRAM, NAND, etc., or may be next-generation semiconductor devices, such as RRAM, STT MRAM, PRAM, etc., or may be a combination thereof.

[0176] When implementing a method (200) for predicting the prognosis of elderly depression patients according to embodiments using an artificial intelligence semiconductor device, the results (weights) of learning a deep learning model using software may be transferred to synapse-mimicking elements arranged in an array, or learning may be performed in the artificial intelligence semiconductor device.

[0177] In the case of implementation by firmware or software, the method (200) for predicting the prognosis of elderly depression patients according to the present embodiments may be implemented in the form of a device, procedure, or function that performs the functions or operations described above. The software code may be stored in a memory unit and executed by a processor. The memory unit may be located inside or outside the processor and may exchange data with the processor by various means already known.

[0178] Additionally, terms such as "system," "processor," "controller," "component," "module," "interface," "model," or "unit" as described above may generally refer to a computer-related entity, such as hardware, a combination of hardware and software, software, or software in execution. For example, the aforementioned components may be, but are not limited to, a process driven by a processor, a processor, a controller, a control processor, an object, a thread of execution, a program, and / or a computer. For example, both an application running on a controller or a processor and the controller or the processor may be components. One or more components may be within a process and / or thread of execution, and the components may be located on a single device (e.g., a system, a computing device, etc.) or distributed across two or more devices.

[0179] Meanwhile, another embodiment provides a computer program stored on a computer storage medium that performs the method (200) for predicting the prognosis of a geriatric patient with depression described above. Furthermore, another embodiment provides a computer-readable storage medium storing a program for implementing the method for analyzing cell pathology slide images described above.

[0180] The program recorded on the recording medium can be read, installed and executed by a computer, thereby executing the steps described above.

[0181] In this way, in order for a computer to read a program recorded on a recording medium and execute functions implemented as a program, the above-mentioned program may include code coded in a computer language such as C, C++, JAVA, or machine language that can be read by the computer's processor (CPU) through the computer's device interface.

[0182] Such code may include functional code related to functions defining the aforementioned functions, and may also include control code related to execution procedures required for the computer's processor to execute the aforementioned functions according to a predetermined procedure.

[0183] Additionally, such code may further include memory reference related code regarding where in the internal or external memory of the computer the additional information or media required for the computer's processor to execute the aforementioned functions should be referenced.

[0184] Additionally, if the computer's processor needs to communicate with another computer or server located remotely in order to execute the functions described above, the code may further include communication-related code regarding how the computer's processor should communicate with another computer or server located remotely using the computer's communication module, and what information or media should be sent and received during the communication.

[0185] The computer-readable recording medium that records the program as described above includes, for example, ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical media storage device, etc., and may also include one implemented in the form of a carrier wave (e.g., transmission via the Internet).

[0186] Additionally, computer-readable recording media can be distributed across network-connected computer systems, allowing computer-readable code to be stored and executed in a distributed manner.

[0187] In addition, the functional program for implementing the present invention and the code and code segments related thereto may be easily inferred or changed by programmers in the technical field to which the present invention belongs, taking into consideration the system environment of the computer that reads the recording medium and executes the program.

[0188] The method (200) for predicting the prognosis of a patient with geriatric depression described through FIG. 16 may also be implemented in the form of a recording medium containing computer-executable instructions, such as an application or program module executed by a computer. The computer-readable medium may be any available medium that can be accessed by a computer, and includes both volatile and nonvolatile media, removable and non-removable media. In addition, the computer-readable medium may include all computer storage media. The computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data.

[0189] The method (200) for predicting the prognosis of a geriatric patient with depression described above can be executed by an application that is installed by default on the terminal (which may include a program included in a platform or operating system installed by default on the terminal), or by an application (i.e., a program) that the user directly installs on the master terminal through an application providing server such as an application store server, an application, or a web server related to the service. In this sense, the method (200) for predicting the prognosis of a geriatric patient described above can be implemented as an application (i.e., a program) that is installed by default on the terminal or directly installed by the user, and can be recorded on a computer-readable recording medium such as on the terminal.

[0190] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single entity may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined manner.

[0191] The scope of the present invention is indicated by the claims described below rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention.

[0192] The above description is merely an illustrative example of the technical idea of ​​the present disclosure, and those skilled in the art to which the present disclosure pertains will appreciate that various modifications and variations can be made without departing from the essential characteristics of the technical idea of ​​the present disclosure. In addition, the present embodiments are not intended to limit the technical idea of ​​the present disclosure but rather to explain it, and therefore the scope of the technical idea of ​​the present disclosure is not limited by these embodiments. The scope of protection of the present disclosure should be interpreted by the claims below, and all technical ideas within a scope equivalent thereto should be interpreted as being included within the scope of the rights of the present disclosure.

[0193]

[0194] CROSS-REFERENCE TO RELATED APPLICATION

[0195] This patent application claims priority under 35 USC §119(a) to Korean Patent Application No. 10-2023-0059432, filed May 8, 2023, the entire contents of which are incorporated herein by reference. Furthermore, this patent application claims priority in countries other than the United States for the same reasons, the entire contents of which are incorporated herein by reference.

Claims

1. A preprocessing unit for preprocessing brain imaging biomarkers of elderly patients with depression; and A device for predicting the prognosis of an elderly patient with depression, comprising a prediction unit for predicting the antidepressant treatment effect of an elderly patient with depression based on a brain imaging biomarker and blood biomarker or a drug administration effect evaluation result of any elderly patient with depression, using a prognosis prediction model for an elderly patient with depression that has been trained to predict the antidepressant treatment effect of the elderly patient with depression using preprocessed brain imaging biomarkers and blood biomarkers or a drug administration effect evaluation result.

2. In paragraph 1, The above brain imaging biomarker is at least one of a high-resolution brain gray matter structural MRI (T1 MRI) image, a T2-Weighted Fluid Attenuated Inversion Recovery (T2 FLAIR) image, and a brain amyloid positron emission tomography (amyloid-PET) image, and a device for predicting the prognosis of an elderly patient with depression.

3. In paragraph 1, A device for predicting the prognosis of an elderly patient with depression, wherein the blood biomarker is at least one of blood oligomerized beta-amyloid (OAB) and blood neurofilament light chain (NF-L).

4. In paragraph 1, The above preprocessing unit preprocesses the image of the brain imaging biomarker by dividing it into specific regions, A device for predicting the prognosis of an elderly patient with depression, wherein the above prediction model is learned to predict the effect of antidepressant treatment of the elderly patient with depression by using the image images of brain imaging biomarkers preprocessed by dividing the region into specific regions by the preprocessing unit and the blood biomarkers or the drug administration effect evaluation results.

5. In paragraph 1, When the above prediction model is trained using the above preprocessed brain imaging biomarkers and blood biomarkers, A device for predicting the prognosis of a patient with geriatric depression, wherein the above prediction model is learned as an initial model for predicting geriatric depression treatment based on the above brain imaging biomarker, and is learned as a model for predicting geriatric depression treatment using the above blood biomarker together with the above brain imaging biomarker.

6. In paragraph 1, When the above prediction model is trained using the preprocessed brain imaging biomarkers and blood biomarkers along with the results of drug administration effect evaluation, If only the results of the drug administration effect evaluation of the above-mentioned arbitrary elderly depression patient are provided, the prediction unit predicts the antidepressant treatment effect of the above-mentioned elderly depression patient for previously used brain imaging biomarkers and blood biomarkers together with the results of the drug administration effect evaluation of the above-mentioned arbitrary elderly depression patient through the prognosis prediction model of the above-mentioned elderly depression patient, A device for predicting the antidepressant treatment effect of an elderly patient with depression for a new brain imaging biomarker and a blood biomarker, together with the results of evaluating the effectiveness of drug administration for an elderly patient with depression, when a new brain imaging biomarker and a blood biomarker are provided together with the results of evaluating the effectiveness of drug administration for an elderly patient with depression, through a prognosis prediction model of the elderly patient with depression.

7. In paragraph 1 The above prediction model includes a first prediction model learned using the preprocessed brain imaging biomarkers and blood biomarkers, and a second prediction model learned using the preprocessed brain imaging biomarkers and blood biomarkers along with the drug administration effect evaluation results. If only the brain imaging biomarkers and blood biomarkers of the above random elderly depression patient are provided, the prediction unit predicts the antidepressant treatment effect of the elderly depression patient through the first prediction model, A device for predicting the antidepressant treatment effect of the elderly patient with depression through the second prediction model, when only the drug administration effect evaluation results are provided or new brain imaging biomarkers and blood biomarkers are provided together with the drug administration effect evaluation results.

8. In paragraph 1, The above prediction unit is a device for predicting the prognosis of an elderly patient with depression, which classifies the elderly patient with depression into one of antidepressant monotherapy, atypical antipsychotic monotherapy, antidepressant + antidepressant treatment, antidepressant + atypical antipsychotic treatment, and antidepressant + basic regulator treatment through a learned prognosis prediction model of an elderly patient with depression.

9. Preprocessing step for preprocessing brain imaging biomarkers of elderly patients with depression; and A method for predicting the prognosis of an elderly patient with depression, comprising a prediction step of predicting the antidepressant treatment effect of an elderly patient with depression for any of the brain imaging biomarkers and blood biomarkers or the results of drug administration effect evaluations of an elderly patient with depression, using a prognosis prediction model of an elderly patient with depression that is trained to predict the antidepressant treatment effect of the elderly patient with depression using the preprocessed brain imaging biomarkers and blood biomarkers or the results of drug administration effect evaluations.

10. In paragraph 9, A method for predicting the prognosis of an elderly patient with depression, wherein the brain imaging biomarker is at least one of a high-resolution brain gray matter structural MRI (T1 MRI) image, a T2-Weighted Fluid Attenuated Inversion Recovery (T2 FLAIR) image, and a brain amyloid positron emission tomography (amyloid-PET) image.

11. In paragraph 9, A method for predicting the prognosis of an elderly patient with depression, wherein the blood biomarker is at least one of blood oligomerized beta-amyloid (OAB) and blood neurofilament light chain (NF-L).

12. In paragraph 9, In the above preprocessing step, the image of the brain imaging biomarker is preprocessed by dividing it into specific regions, A method for predicting the prognosis of an elderly patient with depression, wherein the above prediction model is trained to predict the effect of antidepressant treatment on the elderly patient with depression by using image images of preprocessed brain imaging biomarkers and the blood biomarkers by dividing the region into specific regions through the above preprocessing step.

13. In paragraph 9, When the above prediction model is trained using the above preprocessed brain imaging biomarkers and blood biomarkers, A method for predicting the prognosis of a patient with geriatric depression, wherein the above prediction model is learned as an initial model for predicting geriatric depression treatment based on the above brain imaging biomarker, and the above prediction model is learned as a geriatric depression treatment prediction model using the above blood biomarker together with the above brain imaging biomarker.

14. In paragraph 9, When the above prediction model is trained using the preprocessed brain imaging biomarkers and blood biomarkers along with the results of drug administration effect evaluation, If only the results of the drug administration effect evaluation of the above-mentioned random elderly depression patient are provided, the prognosis prediction model of the above-mentioned random elderly depression patient is used in the above-mentioned prediction step to predict the antidepressant treatment effect of the above-mentioned random elderly depression patient for previously used brain imaging biomarkers and blood biomarkers together with the results of the drug administration effect evaluation of the above-mentioned random elderly depression patient. A method for predicting the antidepressant treatment effect of an elderly patient with depression for a new brain imaging biomarker and a blood biomarker together with the results of evaluating the effectiveness of drug administration for an elderly patient with depression, using a prognosis prediction model for the elderly patient with depression in the prediction step, when new brain imaging biomarkers and blood biomarkers are provided together with the results of evaluating the effectiveness of drug administration for an elderly patient with depression.

15. In paragraph 9 The above prediction model includes a first prediction model learned using the preprocessed brain imaging biomarkers and blood biomarkers, and a second prediction model learned using the preprocessed brain imaging biomarkers and blood biomarkers along with the drug administration effect evaluation results. If only the brain imaging biomarkers and blood biomarkers of the above random elderly depression patient are provided, the antidepressant treatment effect of the elderly depression patient is predicted through the first prediction model in the above prediction step, A method for predicting the antidepressant treatment effect of an elderly patient with depression through the second prediction model in the prediction step, when only the drug administration effect evaluation results are provided or new brain imaging biomarkers and blood biomarkers are provided together with the drug administration effect evaluation results.

16. In paragraph 9, In the above prediction step, A method for predicting the prognosis of an elderly patient with depression, wherein the method classifies the elderly patient with depression into one of antidepressant monotherapy, atypical antipsychotic monotherapy, antidepressant + antidepressant treatment, antidepressant + atypical antipsychotic treatment, and antidepressant + basic modulator treatment using a learned prognosis prediction model of the elderly patient with depression.

17. A memory storing information on preprocessed brain imaging biomarkers and blood biomarkers, information on drug administration effect evaluation results, and a prognosis prediction model for elderly patients with depression that is learned to predict the antidepressant treatment effect of elderly patients with depression using the preprocessed brain imaging biomarkers and blood biomarkers or drug administration effect evaluation results; and A computer device including a processor that, when requested to predict the effect of antidepressant treatment of an elderly patient with depression based on brain imaging biomarkers and blood biomarkers or drug administration effect evaluation results of any elderly patient with depression, preprocesses the brain imaging biomarkers and executes a learned prognosis prediction model of an elderly patient with depression stored in the memory to predict the effect of antidepressant treatment of an elderly patient with depression based on brain imaging biomarkers and blood biomarkers or drug administration effect evaluation results of any elderly patient with depression.

Citation Information

Patent Citations

  • Immediate and long-term lifting cosmetics composite and delay discharge composite mixer of manufacturing the immediate and long-term lifting cosmetics composite

    KR1020210131516A

  • Battery device

    KR1020240077888A

  • System for controlling network access based on controller and method of the same

    KR102333553B1

  • Systems and methods for generating biomarkers based on multivariate classification of functional imaging and associated data

    US20160019693A1

  • Systems and methods for treating mood disorders

    US20210361948A1