Medical information processing device, method, and program

Abstract

To reduce unfairness among subjects regarding allocation of clinical decisions while maintaining the magnitude of benefit across all the subjects.SOLUTION: A medical information processing device according to an embodiment comprises a first acquisition unit, a second acquisition unit, and a determination unit. The first acquisition unit acquires effect evaluation values for multiple clinical decisions. The second acquisition unit acquires fairness indicator values for multiple clinical decisions. The determination unit determines a clinical decision to be allocated to a target subject based on the effect evaluation values and the fairness indicator values.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The embodiments disclosed in the present specification and drawings relate to a medical information processing device, method, and program. [Background technology]

[0002] In clinical research and daily practice, patients are assigned to a specific treatment decision from among multiple treatment decisions. Methods for adaptively changing allocation include adaptive design and bandit algorithms. To maximize the benefit to all patients, it is necessary to deliberately select treatment decisions that are unlikely to be optimal at an appropriate frequency. However, because the estimated magnitude of the effect of each treatment decision is known at the time of allocation, unfairness between patients occurs. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2021 / 065845 Summary of the Invention [Problem to be solved by the invention]

[0004] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to reduce unfairness among subjects in the allocation of clinical decisions while maintaining the magnitude of benefit to all subjects. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described below can also be positioned as other problems. [Means for solving the problem]

[0005] A medical information processing apparatus according to an embodiment includes a first acquisition unit, a second acquisition unit, and a determination unit. The first acquisition unit acquires effect evaluation values ​​for a plurality of medical decisions. The second acquisition unit acquires fairness index values ​​for the plurality of medical decisions. The determination unit determines a medical decision to be assigned to a target subject based on the effect evaluation values ​​and the fairness index values. [Brief explanation of the drawings]

[0006] [Figure 1] FIG. 1 is a diagram showing an example of the configuration of a medical image processing apparatus according to this embodiment. [Figure 2] FIG. 2 is a diagram showing the flow of medical information processing performed by a processing circuit in accordance with a medical information processing program. [Figure 3] FIG. 3 is a diagram schematically illustrating the medical information processing shown in FIG. [Figure 4] FIG. 4 is a diagram illustrating a flow of the allocation process performed by the processing circuit according to the first embodiment. [Figure 5] FIG. 5 is a diagram schematically illustrating the medical information processing shown in FIG. [Figure 6] FIG. 6 is a diagram illustrating the concept of the subjective optimal arm probability and the subjective optimal arm probability difference. [Figure 7] FIG. 7 is a diagram illustrating fairness index values ​​when the target value is within a range. [Figure 8] FIG. 8 is a diagram illustrating a flow of the allocation process performed by the processing circuit according to the second embodiment. [Figure 9] FIG. 9 is a diagram schematically illustrating the medical information processing shown in FIG. [Figure 10] FIG. 10 is a diagram illustrating the fairness index value when the target value is a single value. [Figure 11] FIG. 11 is a diagram showing the transition of the allocation policy for medical care decisions based on a single target value. [Figure 12] FIG. 12 shows a specific example of assignment setting processing. [Figure 13] FIG. 13 shows another specific example of the assignment setting process. DETAILED DESCRIPTION OF THE INVENTION

[0007] Hereinafter, embodiments of a medical information processing device, method, and program will be described in detail with reference to the drawings.

[0008] In the medical field, efforts are being made to realize clinical decision support (CDS) using data and artificial intelligence (AI) technology. Supervised learning, typified by deep learning, is widely used in AI technology, and some AI technologies have emerged that can detect lesions with greater accuracy than humans, particularly in the field of diagnostic imaging.

[0009] However, the following issues may arise when implementing clinical decision support using supervised learning. 1. In order to improve accuracy, a large amount of data is required, and the cost of collecting the data is high. 2. Even if the accuracy is high, it may not lead to appropriate medical decision-making.

[0010] Problem 2 is technically equivalent to the problem that "supervised learning can reveal correlations, but not causal relationships." For example, even if there is an AI model that can predict with 100% accuracy whether a patient with heart failure will die within five years, this AI model will be useless to a doctor who wants to know how to extend this patient's life expectancy to more than five years.

[0011] Identifying causal relationships in medical treatment and identifying optimal treatment is something that cannot currently be achieved through supervised learning; this is the role of randomized controlled trials (RCTs) (or medical statistics). In other words, with the extension of current AI technology, there is no true "clinical decision support for treatment optimization," and randomized controlled trials will never disappear.

[0012] However, randomized controlled trials also have many problems, such as the following: 1. Because it takes a long time to complete research, it is difficult to keep up with the latest trends, and the results are likely to become outdated. 2. Research is inefficient because it requires huge costs and there is a risk of hypothesis verification failing. 3. Eligibility criteria for studies tend to be strict, so the results may not apply to the general patient population (low external validity). 4. Only the average effect of interventions can be compared, and it is not possible to clarify the different effects for each patient. For example, it is not possible to consider cases such as personalized medicine. 5. Overreliance on statistical significance is prone to publication bias and p-value hacking.

[0013] In particular, the limitations of randomized controlled trials are becoming apparent these days as treatment methods change rapidly and the importance of personalized medicine is emphasized.

[0014] Therefore, there is a need to efficiently optimize clinical decisions without requiring large amounts of prior data or conducting randomized controlled trials, and there is also a need to develop a clinical decision support model for this purpose.

[0015] Fig. 1 is a diagram showing an example of the configuration of a medical information processing device 1 according to this embodiment. As shown in Fig. 1, the medical information processing device 1 is an information processing terminal such as a computer having a processing circuit 11, a storage device 12, an input device 13, a communication device 14, and a display device 15. The processing circuit 11, the storage device 12, the input device 13, the communication device 14, and the display device 15 are connected via a bus so that signals can be input and output to and from each other.

[0016] The processing circuitry 11 has processors such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit). The processing circuitry 11 executes a medical information processing program to realize an allocation function 111, an observation function 112, a storage function 113, an update function 114, a display control function 115, and the like. Note that each of the functions 111 to 115 is not limited to being realized by a single processing circuit. A processing circuit may be configured by combining multiple independent processors, and each processor may execute a program to realize each of the functions 111 to 115. Furthermore, the functions 111 to 115 may each be a modular program that constitutes a medical information processing program. These programs are stored in the storage device 12.

[0017] The storage device 12 is a read-only memory (ROM), a random access memory (RAM), a hard disk drive (HDD), a solid state drive (SSD), an integrated circuit storage device, or the like that stores various information. In addition to the above storage devices, the storage device 12 may also be a portable recording medium such as a compact disc (CD), a digital versatile disc (DVD), or a flash memory, or a drive that reads and writes various information from and to a semiconductor memory element, or the like. The storage device 12 may also be located in another computer connected via a network.

[0018] The input device 13 receives various input operations from an operator, converts the received input operations into electrical signals, and outputs the electrical signals to the processing circuit 11. Specifically, the input device 13 is connected to input devices such as a mouse, a keyboard, a trackball, a switch, a button, a joystick, a touchpad, and a touch panel display. The input device 13 outputs electrical signals corresponding to the input operations to the input device to the processing circuit 11. The input device 13 may also be an input device provided in another computer connected via a network or the like.

[0019] The communication device 14 is an interface for transmitting and receiving various types of information to and from other computers. Information communication by the communication device 14 is performed in accordance with a standard suitable for medical information communication, such as DICOM (Digital Imaging and Communications in Medicine).

[0020] The display device 15 displays various information using the display control function 115 of the processing circuit 11. As the display device 15, for example, a liquid crystal display (LCD), a cathode ray tube (CRT) display, an organic electroluminescence display (OLED), a plasma display, or any other display can be used as appropriate. Also, a projector may be used as the display device 15.

[0021] The processing circuit 11 according to this embodiment determines a medical judgment to be assigned to a subject, such as a patient, based on the effect evaluation value of the medical judgment by implementing an allocation function 111. The effect evaluation value of the medical judgment is calculated using a model for calculating the effect evaluation value of the medical judgment. A medical procedure corresponding to the assigned medical judgment is administered to the subject. An effect resulting from the medical procedure occurs in the subject. The processing circuit 11 observes the effect occurring in the subject by implementing an observation function 112. The effect is expressed as a numerical value, which will be referred to as an effect observation value. The processing circuit 11 associates the effect observation value, the medical procedure, and other features and stores them as observation data in the storage device 12 by implementing a storage function 113. The above-described medical judgment assignment, effect observation, and observation data storage are performed for various subjects. The processing circuit 11 updates the model based on the observation data when the observation data is stored by implementing an update function 114. The processing circuit 11 displays various data on the display device 15 by implementing a display control function 115.

[0022] The allocation function 111 is divided into an effect evaluation value acquisition function 116, a fairness index value acquisition function 117, and an allocation determination function 118. By realizing the effect evaluation value acquisition function 116, the processing circuit 11 acquires effect evaluation values ​​for multiple medical decisions. By realizing the fairness index value acquisition function 117, the processing circuit 11 acquires fairness index values ​​for multiple medical decisions. By realizing the allocation determination function 118, the processing circuit 11 determines a medical decision to be allocated to a target subject based on the effect evaluation value and the fairness index value.

[0023] The subject according to this embodiment is typically a single patient with a disease. However, the subject according to this embodiment is not limited to this and may be a patient group consisting of two or more patients. Furthermore, the subject according to this embodiment does not necessarily have to be a person with a disease, but may also be a healthy person. The model may be used for each doctor, each department, each hospital, each region, or each subject.

[0024] The model according to this embodiment is a function for calculating the clinical decision-making effect evaluation value, which may be a function determined manually, a function determined by an experimental method or a deterministic method, a function trained by machine learning, or a function determined by any other method.

[0025] Types of clinical decisions related to this embodiment are expected to include decisions regarding whether to perform a procedure, decisions regarding content, decisions regarding amount, and decisions regarding timing. Examples of decisions regarding whether to perform a procedure include decisions regarding whether to perform surgery or whether to perform a blood test. Examples of decisions regarding content include decisions regarding the selection of a therapeutic drug to be used or the selection of a disease name to diagnose. Examples of decisions regarding amount include decisions regarding the selection of a drug dosage or the selection of rehabilitation implementation time. Examples of decisions regarding timing include decisions regarding the selection of when to visit the hospital or when to perform surgery.

[0026] Those who make medical decisions include not only medical professionals such as doctors, nurses, and paramedical staff, but also the subjects themselves, their families, and anyone related to the subject. Medical decisions do not necessarily have to be highly medical, nor do they have to be decisions made for the purpose of improving health. In other words, decisions that have a negative effect on medical care or health also fall within the scope of medical decisions. For example, the decision of whether or not a healthy person should smoke at a certain time is also within the scope of medical decisions.

[0027] As a result of a clinical decision, some effect occurs. This effect is also called a reward or an outcome. Possible effects according to this embodiment include, for example, clinical outcomes, patient-reported outcomes, and economic outcomes. Examples of clinical outcomes include incidence rates (including the presence or absence of morbidity), 5-year survival rates (including the presence or absence of survival), complication rates (including the presence or absence of complications), readmission rates (including the presence or absence of readmission), test values ​​(or the degree of improvement in test values), and the degree of independence in daily living. Examples of patient-reported outcomes include subjective symptoms, subjective health status, satisfaction with treatment, and subjective happiness. Examples of economic outcomes include medical expenses, invested medical resources, and length of hospital stay.

[0028] Effects are expressed as numerical values ​​and can be ranked for learning purposes. Numeric values ​​may be assigned to things that are not originally numerical. Effects may be observed immediately after a clinical decision. For example, the effect on a clinical decision on whether to notify a healthy individual's smartphone of a message encouraging exercise may be whether or not the individual exercises within five minutes of receiving the message.

[0029] The effectiveness can also take into account the cost required for medical decisions. For example, in the smartphone message example above, the benefit of receiving a notification is that it is easier to achieve the desired effect than not receiving a notification. However, notifications also have the disadvantage of increasing communication costs and hindering user behavior. To take these advantages and disadvantages into account, for example, if the first message always incurs a cost of "5" and the exerciser receives a reward of "100," the effectiveness can be calculated as follows: Reward "100" - Cost "5" = Effectiveness "95." This makes it possible to obtain an effect that reflects cost-effectiveness.

[0030] The feature quantities according to this embodiment include the attributes and / or conditions of the subject. The attributes are information such as the gender and age of the subject, which have properties that do not change depending on the immediately preceding medical decision. The conditions are information such as the blood pressure and blood glucose levels of the subject, which have properties that change depending on the immediately preceding medical decision.

[0031] An example of the operation of the medical information processing apparatus 1 according to this embodiment will be described below.

[0032] 2 is a diagram showing the flow of medical information processing performed by the processing circuitry 11 in accordance with the medical information processing program. FIG. 3 is a diagram showing a schematic diagram of the medical information processing shown in FIG.

[0033] As shown in Figures 2 and 3, the processing circuitry 11, by implementing the allocation function 111, allocates a medical decision to a subject using a model for the subject (step S1). In step S1, the processing circuitry 11 determines a medical decision to be assigned according to an algorithm based on a bandit algorithm. The bandit algorithm is an algorithm that aims to optimize the trade-off between exploration and exploitation, and sequentially determines a medical decision to maximize the cumulative reward or minimize the cumulative regret (regret represents the difference between the reward obtained from the optimal action and the reward obtained from the actual action). Even if the cumulative reward is maximized or the cumulative regret is minimized, fairness of the search is not necessarily ensured. Therefore, to maximize the cumulative reward or minimize the cumulative regret while maintaining fairness of the search, the processing circuitry 11 determines a medical decision to be assigned based on the effect evaluation value and the fairness index value. More specifically, in step S1, medical decisions are allocated according to the following process.

[0034] The bandit algorithm according to this embodiment includes not only a narrow bandit algorithm whose effect does not depend on features, but also a contextual bandit algorithm whose effect depends on features, and reinforcement learning that solves a sequential decision-making problem in which the state changes depending on previous clinical decisions. Specific examples of the bandit algorithm according to this embodiment include Epsilon Greedy, Thompson Sampling, Linear Thompson Sampling, PSRL (Posterior Sampling for Reinforcement Learning), BDQN (Bayesian Deep Q-Networks), etc.

[0035] First, the processing circuitry 11 calculates a plurality of effect evaluation values ​​corresponding to a plurality of medical decisions using a model by implementing the effect evaluation value acquisition function 116 (step S101). The effect evaluation value is an index value for evaluating the effect of the medical decisions of the allocation candidates, and may be, for example, an index value such as the subjective optimal arm probability, the subjective optimal arm probability difference, the confidence interval of the effect, the expected value of the effect, the maximum value of the expected value, or the difference from the maximum value of the expected value.

[0036] When step S101 is performed, the processing circuit 11 calculates a plurality of fairness index values ​​corresponding to a plurality of medical decisions, respectively, by implementing the fairness index value acquisition function 117 (step S102). A fairness index value is calculated for each medical decision. The fairness index value is an index value that evaluates the degree of fairness of selecting the medical decision. The fairness index value is an index value based on the effectiveness evaluation value. More specifically, the fairness index value is calculated based on the effectiveness evaluation value and its target value. The fairness index value according to this embodiment may be either a range or a single value as the target value.

[0037] When step S102 is performed, the processing circuitry 11, by realizing the allocation determination function 118, determines a medical decision to be assigned to a subject based on the effect evaluation value calculated in step S101 and the fairness index value calculated in step S102 (step S103). Specifically, the processing circuitry 11 according to the following Example 1 determines a single random number according to a predetermined probability distribution, and determines one or more allocation candidates from among multiple medical decisions according to a predetermined strategy based on the determined single random number. If the fairness index value for an allocation candidate satisfies a predetermined condition, the processing circuitry 11 determines the allocation candidate as the medical decision to be assigned. The processing circuitry 11 according to the following Example 2 determines multiple random numbers from among multiple medical decisions according to a predetermined probability distribution, and determines multiple allocation candidates probabilistically according to a predetermined strategy based on the determined multiple random numbers. The processing circuitry 11 determines the medical decision to be assigned based on a comparison of the fairness index values ​​among the multiple allocation candidates.

[0038] When step S103 is performed, a medical procedure corresponding to the medical judgment is performed on the subject. At this time, the medical procedure may be performed corresponding to a medical judgment selected by the subject himself, rather than the medical judgment determined in step S103.

[0039] When step S1 is performed, the processing circuit 11 observes the effect occurring in the subject by implementing the observation function 112 (step S2). The effect is observed numerically as an effect observation value. For example, if the clinical decision is "perform surgery," the effect observation value may be a 5-year survival rate or the occurrence of complications. The effect observation value may be obtained by any method. For example, it may be input by an operator via the input device 13, or a measurement value obtained by an examination device may be input. Alternatively, it may be received from another computer via the communication device 14.

[0040] When step S2 is performed, the processing circuit 11 realizes the storage function 113 to store the observation data in a database DB managed by the storage device 12 (step S3). The observation data includes the subject's identifier, a medical judgment corresponding to the medical procedure performed on the subject, and an effect observation value of the effect produced in the subject by the medical procedure.

[0041] After step S3 is performed, the processing circuit 11 updates the model by implementing the update function 114 (step S4). As an example, the processing circuit 11 updates the parameters of the model every time an effect is observed, in other words, every time an effect observation value is obtained.

[0042] This completes the medical information processing according to this embodiment. Steps S1 to S4 are then repeatedly executed for the same subject or different subjects. This improves the accuracy of calculation of the model's effect evaluation value and also improves the accuracy of allocation of clinical decisions.

[0043] Several examples of this embodiment will be described below.

[0044] Example 1 Bandit algorithms, including Thompson sampling, are theoretically guaranteed to draw options that do not maximize expected value the "minimum necessary" number of times. However, this "minimum necessary" refers to the selection of all options, including an infinite number of subjects. With a finite number of subjects, unnecessary exploration is likely to occur. For example, the benefits of selecting an option that is highly unlikely to maximize expected value for a given subject will occur far into the future. On the other hand, such exploration significantly reduces fairness between subjects, since it selects an option that is highly unlikely to maximize expected value. Thus, when a finite number of explorations is assumed, eliminating some explorations may improve both regret and fairness.

[0045] The processing circuit 11 according to the first embodiment sets a threshold value for the effect evaluation value and executes a technique (clipping) to prevent medical decisions having an effect evaluation value greater than the threshold value from being selected. That is, in the first embodiment, a target value for the effect evaluation value is set within a range. The medical information processing according to the first embodiment will be described below. The effect evaluation value according to the first embodiment is assumed to be a subjective optimal arm probability and / or a subjective optimal arm probability difference.

[0046] Fig. 4 is a diagram showing the flow of allocation processing by the processing circuit 11 according to the first embodiment. Fig. 5 is a diagram schematically showing the medical information processing shown in Fig. 4. Steps S111 to S118 shown in Fig. 4 are included in step S1 shown in Fig. 2. The medical information processing shown in Fig. 4 is assumed to be the n-th allocation step among the allocation steps (S1) repeatedly performed in Fig. 2. In other words, it is assumed that allocation of medical decisions has already been performed multiple times for the same or different subjects.

[0047] First, the processing circuit 11 calculates the subjective optimal arm probability for each medical decision by implementing the effect evaluation value acquisition function 116 (step S111). The subjective optimal arm probability is the subjective probability of the event that the expected value of the effect of that medical decision is the highest at that time. The subjective probability refers to the degree of subjective belief or certainty based on the information and experience possessed by an individual or algorithm. The subjective optimal arm probability is calculated based on the posterior distribution of the effect for each medical decision, according to the following formula (1) and based on Bayes' theorem. Here, q i (α|D i ) is the observed data (effect observed value) D i is the posterior probability density function of the effect of clinical decision i under the obtained conditions, and α and β are parameters of the probability density function. The subjective optimal arm probability can be calculated using Monte Carlo simulation, etc.

[0048]

number

[0049] When step S111 is performed, the processing circuitry 11, by implementing the allocation determination function 118, determines allocation candidates based on the subjective optimal arm probability according to a bandit algorithm (step S112). The specific strategy algorithm of the bandit algorithm is not particularly limited, and examples of the algorithm that can be used include the Epsilon Greedy method, the Upper Confidence Bound (UCB) method, the Minimum Empirical Divergence (MED) method, and Thompson sampling. For example, when Thompson sampling is used, the processing circuitry 11 determines medical decisions probabilistically so that the allocation probability (sampling probability) of each medical decision matches the subjective optimal arm probability. The determined medical decisions are set as allocation candidates.

[0050] As shown in Fig. 5, in the first embodiment, a medical decision to be allocated is determined from six medical decisions A to F. In step S111, the processing circuitry 11 first samples a single random number from an arbitrary probability distribution. Then, the processing circuitry 11 determines one allocation candidate from the six medical decisions A to F according to a predetermined policy algorithm based on the single random number. The predetermined policy algorithm is not limited to a probabilistic policy, but may be a deterministic policy, and can be arbitrarily selected from the above policy algorithms. In Fig. 5, it is assumed that medical decision C is determined as an allocation candidate.

[0051] When step S112 is performed, the processing circuit 11 realizes the effectiveness evaluation value acquisition function 116 to calculate the subjective optimal arm probability difference for the allocation candidate (step S113). The subjective optimal arm probability is an example of an effectiveness evaluation value. The subjective optimal arm probability difference Δp i As expressed in the following equation (2), the maximum value of the subjective optimal arm probability of all medical decisions is max j p j From the above, the subjective optimal arm probability p j This is the value obtained by subtracting

[0052]

number

[0053] Figure 6 is a diagram illustrating the concepts of subjective optimal arm probability and subjective optimal arm probability difference. As shown in Figure 6, the posterior distribution a1 of the treatment effect for medical decision 1 and the posterior distribution a2 of the treatment effect for medical decision 2 are illustrated. Note that the treatment effect is an example of the effect of a medical decision. If the subjective optimal arm probability for medical decision 1 is 10% and the subjective optimal arm probability for medical decision 2 is 90%, then the maximum value of the subjective optimal arm probabilities between medical decisions 1 and 2 is 90%, so the subjective optimal arm probability difference for medical decision 1 is 90% - 10% = 80%, and the subjective optimal arm probability difference for medical decision 2 is 90% - 90% = 0%.

[0054] After step S113 is performed, the processing circuit 11 sets a target value for the subjective optimal arm probability difference by implementing the fairness index value acquisition function 117 (step S114). The target value according to the first embodiment is set within the range of the subjective optimal arm probability difference. After step S114 is performed, the processing circuit 11 calculates a fairness index value based on the subjective optimal arm probability difference and the target value by implementing the fairness index value acquisition function 117 (step S115).

[0055] FIG. 7 is a diagram showing the fairness index value when the target value is a range. As shown in FIG. 7, a threshold is set to an arbitrary value of the subjective optimal arm probability difference, which is the effectiveness evaluation value. A range of the subjective optimal arm probability difference below the threshold is set as the target value. A fairness index value of "1" is assigned to the range from the threshold to a subjective optimal arm probability difference of 100%. A fairness index value of "1" means unfairness. A fairness index value of "0" is assigned to the range from the threshold to a subjective optimal arm probability difference of 0%, i.e., the range of the target value. A fairness index value of "0" means fairness.

[0056] For example, consider a case where the subjective optimal arm probability difference for medical judgment A is 0% and the subjective optimal arm probability difference for medical judgment B is 80%, and the target value is fixed to the range of 0% to 40%. In this case, since the subjective optimal arm probability difference for medical judgment A is within the target value range, the fairness index value is set to "0", indicating fairness. On the other hand, since the subjective optimal arm probability difference for medical judgment B is outside the target value range, the fairness index value is set to "1", indicating unfairness.

[0057] The target value may or may not be a fixed value across multiple processing times (trials). The target value may also vary depending on the number of trials. For example, the target value may be set to narrower as the number of trials increases, i.e., the range of fairness may become narrower.

[0058] When step S115 is performed, the processing circuit 11, by realizing the allocation determination function 118, determines whether or not a fairness index value of "0" was set to the allocation candidate in step S115 (step S116). If it is determined in step S116 that a fairness index value of "0" was set to the allocation candidate (step S116: YES), the processing circuit 11 sets the allocation candidate as the medical judgment to be assigned (step S117). If it is not determined in step S116 that a fairness index value of "0" was set to the allocation candidate (step S116: NO), that is, if a fairness index value of "1" was set to the allocation candidate, the processing circuit 11 sets another medical judgment as the medical judgment to be assigned (step S118).

[0059] For example, as shown in FIG. 5, in step S116, it is determined whether a fairness index value of "0" has been assigned to medical judgment C, which is an allocation candidate. If it is determined that a fairness index value of "0" has been assigned to medical judgment C, medical judgment C is adopted for allocation. On the other hand, if it is determined that a fairness index value of "1" has been assigned to medical judgment C, medical judgment C is rejected. In this case, the processing circuitry 11 may assign the medical judgment with the highest subjective optimal arm probability among medical judgments A, B, D, E, and F other than the allocation candidates to the allocation candidate. As another example, the processing circuitry 11 may probabilistically assign the medical judgment with a fairness index value of "0" among medical judgments A, B, D, E, and F other than the allocation candidates. In FIG. 5, medical judgment F is assigned.

[0060] This completes the allocation process according to the first embodiment.

[0061] As described above, the processing circuit 11 according to the first embodiment determines a single random number according to a predetermined probability distribution, and then determines one or more allocation candidates from among multiple medical decisions according to a predetermined strategy based on the determined single random number. If the fairness index value for the allocation candidate satisfies a predetermined condition, the processing circuit 11 determines the allocation candidate as the medical decision to be assigned. More specifically, if the fairness index value of the allocation candidate is a predetermined value "0," the processing circuit 11 assigns the allocation candidate to the medical decision to be assigned. If the fairness index value of the allocation candidate is not the predetermined value "0," the processing circuit 11 assigns a medical decision other than the allocation candidate to the medical decision to be assigned. The method according to the first embodiment can be said to correct the allocation selected by the bandit algorithm based on the fairness index value. The method according to the first embodiment is based on a bandit algorithm that selects medical decisions at an appropriate search frequency, and discards allocation candidates that lack fairness from the allocation candidates, thereby enabling search while ensuring fairness.

[0062] The method according to Example 1 is applicable to any of the following decision-making problems: context-free bandit, contextual bandit, and sequential decision-making, because the fairness index only affects the selection of a clinical decision. The context-free bandit is a case in which the effect depends only on the clinical decision. When each subject selects one of multiple clinical decisions, the expected effect of the subject is affected only by the selected clinical decision. This type of problem is generally called a multi-armed bandit. The contextual bandit is a case in which the effect depends not only on the clinical decision but also on the patient's characteristics. Even with the same clinical decision, the effect of the clinical decision varies depending on the subject. In such a case, optimizing the clinical decision for each individual is equivalent to realizing precision medicine. Sequential decision-making is a case in which each subject makes multiple choices, and the choices are not independent. For example, consider the case in which a diabetic patient is periodically administered an arbitrary amount of insulin (a hormone drug that lowers blood sugar levels) to control their blood sugar levels. In this case, the blood glucose level obtained as a result of this insulin administration depends on the blood glucose level before administration, and this blood glucose level before administration depends on the result of the previous insulin administration. In this way, sequential decision-making is a problem in which the next decision is made repeatedly based on the effects observed as a result of one decision and new features. In medicine, this is equivalent to dynamically optimizing clinical pathways (which define the sequence and implementation standards of treatments and tests) and chronic care management decisions.

[0063] Example 2 In a method based on Thompson sampling, because clinical decisions are determined probabilistically, even if a target value for the subjective optimal arm probability difference is set, it is essentially impossible to approach that target value. In Example 2, we consider approaching the target value for the subjective optimal arm probability difference by allowing a certain degree of leeway in the selection of clinical decisions.

[0064] The processing circuit 11 according to the second embodiment executes a method (targeting) that combines the above-described target value setting and action selection method in sequential decision-making. In the second embodiment, the target value for the subjective optimal arm probability difference is set as a single value. Hereinafter, medical information processing according to the second embodiment will be described. As in the first embodiment, the effect evaluation value according to the second embodiment is also assumed to be the subjective optimal arm probability and / or the subjective optimal arm probability difference.

[0065] Fig. 8 is a diagram showing the flow of allocation processing by the processing circuitry 11 according to the second embodiment. Fig. 9 is a diagram schematically showing the medical information processing shown in Fig. 8. Steps S121 to S126 shown in Fig. 8 are included in step S1 shown in Fig. 2. The medical information processing shown in Fig. 8 is assumed to be the n-th allocation step among the allocation steps (S1) repeatedly performed in Fig. 2. In other words, it is assumed that allocation of medical decisions has already been performed multiple times for the same or different subjects.

[0066] First, the processing circuitry 11 calculates the subjective optimal arm probability for each medical decision by implementing the effect evaluation value acquisition function 116 (step S121). Step S121 is the same as step S111.

[0067] When step S121 is performed, the processing circuitry 11, by implementing the allocation determination function 118, determines multiple allocation candidates based on the subjective optimal arm probability in accordance with a bandit algorithm (step S122). In step S122, the processing circuitry 11 first independently samples multiple random numbers from an arbitrary distribution. For example, as shown in Figure 9, consider the case where an allocation target is determined from medical judgment A and medical judgment B. First, multiple random numbers (five in Figure 9) are sampled from an arbitrary probability distribution. Then, the processing circuitry 11 determines an allocation candidate from medical judgment A and medical judgment B in accordance with a predetermined algorithm based on each of the five random numbers. Note that the predetermined algorithm is not limited to a probabilistic strategy and may be a deterministic strategy, and can be arbitrarily selected from the above-mentioned strategy algorithms.

[0068] After step S122, the processing circuit 11 calculates the subjective optimal arm probability difference for each allocation candidate (step S123) by implementing the effect evaluation value acquisition function 116. The method for calculating the subjective optimal arm probability difference is the same as in the first embodiment.

[0069] When step S123 is performed, the processing circuit 11 sets a target value for the subjective optimal arm probability difference by implementing the fairness index value acquisition function 117 (step S124). The target value according to the second embodiment is set to a single value. When step S124 is performed, the processing circuit 11 calculates a fairness index value based on the subjective optimal arm probability difference and the target value by implementing the fairness index value acquisition function 117 (step S125).

[0070] Fig. 10 is a diagram showing a fairness index value when the target value is a single value. As shown in Fig. 10, in Example 2, the target value is set to a single value of the subjective optimal arm probability difference. The target value is set to a fairness index value of "1", which indicates fairness. The fairness index value is set continuously or discretely so that the farther it is from the target value, the closer it is to "0", which indicates unfairness.

[0071] The processing circuit 11 according to the second embodiment calculates a target value based on the "effect evaluation value of the target subject's past medical decisions" and / or the "effect evaluation value of the medical decisions selected by one or more subjects other than the target subject." That is, the target value may change depending on the medical decisions selected by other subjects, or may change depending on the medical decisions selected by the same subject up to the last time.

[0072] As an example, the processing circuit 11 calculates the target value Δp based on the exponential moving average of the first derivative of the within-subject average of the subjective optimal arm probability difference in accordance with the following equation (3): target Update Δp e is the average subjective optimal arm probability difference for subject e, and Δp e-1is the average value of the subjective optimal arm probability difference for subject e-1. Subject e-1 is the person who was assigned to the treatment decision immediately before subject e. In equation (3), β is a hyperparameter for calculating momentum. According to equation (3), the target value Δp target is the average subjective optimal arm probability difference Δp e and the mean difference in subjective optimal arm probability for subject e-1, Δp e-1 and is updated based on the

[0073]

number

[0074] Figure 11 shows the transition of the allocation policy for medical decisions based on a single target value. In Figure 11, patient 3 is the subject target for allocation of medical decisions, and the stage is to determine the allocation of the second medical decision I1. As shown in the left diagram of Figure 11, for patient 1, the subjective optimal arm probability difference for the first medical decision is "10%", the subjective optimal arm probability difference for the second medical decision is "0%," and the subjective optimal arm probability difference for the third medical decision is "20%". The average value of the subjective optimal arm probability difference for patient 1, Δp e-1 is "10%". For patient 2, the subjective optimal arm probability difference for the first medical decision is "30%", the subjective optimal arm probability difference for the second medical decision is "10%," and the subjective optimal arm probability difference for the third medical decision is "20%". The average subjective optimal arm probability difference for patient 2, Δp e is "20%". For patient 3, the subjective optimal arm probability difference for the first clinical decision is "0%", and the average value of the subjective optimal arm probability difference for patient 3 is "0%". In this case, according to the above formula (3), if the hyperparameter β is 0, the patient's target value Δp target becomes "30%". The model aims to select a treatment decision with a subjective optimal arm probability difference of "60%" as the second treatment decision I1 for patient 3. However, there may be cases where a treatment decision with a subjective optimal arm probability difference of "60%" does not exist, and even if such a treatment decision exists, it is not necessarily the case that patient B will actually select that treatment decision.

[0075] As shown in the right diagram of Figure 11, assume that patient 3 selects a treatment decision with a subjective optimal arm probability difference of 40% for his / her second treatment decision. In this case, the average subjective optimal arm probability difference for patient 3 is (0% + 40%) / 2 = 20%. The model aims to select a treatment decision with a subjective optimal arm probability difference of 50% for patient 3's third treatment decision I2 so that the average subjective optimal arm probability difference of the three treatment decisions matches the target value. In this way, in a sequential decision-making problem, a single target value is set, and the allocation policy for the current treatment decision changes depending on the subjective optimal arm probability difference of the patient's previous treatment decisions. This allows treatment decision allocation to be adaptive to the patient's most recent selection.

[0076] As in the above example, the target value may be set not only based on the average value of the subjective optimal arm probability differences of the other subjects, but also based on only the average value of the subject's own subjective optimal arm probability difference, or on both the average value of the subject's own subjective optimal arm probability difference and the average value of the subjective optimal arm probability differences of the other subjects.

[0077] When step S125 is performed, the processing circuit 11, by implementing the allocation determination function 118, sets a medical decision to be allocated from among the allocation candidates based on the fairness index value (step S126).

[0078] FIG. 12 shows a specific example of the allocation candidate setting process. As shown in FIG. 12, assume that the subjective optimal arm probability of medical judgment A is "60%" and the subjective optimal arm probability difference is "0%," and the subjective optimal arm probability of medical judgment B is "40%" and the subjective optimal arm probability difference is "20%." In step S122, medical judgment A and medical judgment B are probabilistically sampled the number of times equal to the number of random numbers, with an allocation probability (sampling probability) corresponding to the subjective optimal arm probability. In the example of FIG. 12, if there are five random numbers, medical judgment A and medical judgment B are sampled five times at a ratio of 6:4, resulting in five allocation candidates. In step S126, the allocation candidate with the best fairness index value is set as the allocation candidate.

[0079] Here, if the target average subjective optimal arm probability difference is "10%" and the average of the subjective optimal arm probability differences for the past two selections is "0%, " then in order for the average of the subjective optimal arm probability differences for the three selections, including the past two selections and the current selection, to approach the target average of "10%,," it would be ideal to assign a clinical decision with a subjective optimal arm probability difference of "30%" to each allocation candidate. In this case, the target value is set to "30%." The processing circuit 11 calculates a fairness index value for each allocation candidate based on the difference between the target value and the subjective optimal arm probability difference. The fairness index value ranges, for example, from a maximum value of "1" indicating unfairness to a minimum value of "0" indicating fairness, and is set so that it approaches the minimum value of "0" as the difference value decreases. Since allocation candidate A has a subjective optimal arm probability difference of "0%," the fairness index value is set to "0.3", and since allocation candidate B has a subjective optimal arm probability difference of "20%," the fairness index value is set to "0.1". Therefore, since allocation candidate B has a smaller fairness index value than allocation candidate A, allocation candidate B is set as the allocation target.

[0080] Figure 13 shows another specific example of the allocation target selection process. As shown in Figure 13, assume that medical decision A has a subjective optimal arm probability of 90% and a subjective optimal arm probability difference of 0%, while medical decision B has a subjective optimal arm probability of 10% and a subjective optimal arm probability difference of 80%. In the example of Figure 13, medical decisions A and B are sampled five times at a ratio of 9:1, resulting in five allocation candidates. Here, if the target average subjective optimal arm probability difference is 40% and the average of the subjective optimal arm probability differences over the past two selections is 0%, ideally the medical decision with a subjective optimal arm probability difference of 120% should be selected as the allocation target, since the average of the subjective optimal arm probability differences over the three selections will approach the target average of 40%. Here, of medical decisions A and B, medical decision B has the subjective optimal arm probability difference closest to "120%", but since medical decision B has not been sampled as an allocation candidate, the processing circuit 11 sets medical decision A as the allocation target.

[0081] This completes the allocation process according to the second embodiment.

[0082] As described above, the processing circuit 11 according to the second embodiment determines multiple random numbers from multiple clinical decisions according to a predetermined probability distribution, and probabilistically determines multiple allocation candidates based on the determined random numbers according to a predetermined strategy. The processing circuit 11 determines the clinical decision to be assigned based on a comparison of fairness index values ​​among the multiple allocation candidates. More specifically, the processing circuit 11 assigns the allocation candidate with the best fairness index value among the multiple allocation candidates as the allocation candidate. The method according to the second embodiment is based on a bandit algorithm in which clinical decisions are determined probabilistically, but by sampling multiple allocation candidates, it is possible to increase the likelihood of selecting a clinical decision close to the target value of the subjective optimal arm probability difference. Furthermore, the method according to the second embodiment calculates the target value based on the "past effect evaluation value of the target subject" and / or the "effect evaluation value of clinical decisions selected by one or more subjects other than the target subject," making it applicable to sequential decision-making problems.

[0083] (others) The above-mentioned Examples 1 and 2 are examples of medical information processing according to this embodiment, and the steps of the medical information processing according to this embodiment are not limited to these. The medical information processing device according to this embodiment includes any process that determines the medical treatment decision for the allocation target based on the effect evaluation value and the fairness index value.

[0084] In some of the above-described embodiments, the processing circuitry 11 calculates the effect evaluation value by implementing the effect evaluation value acquisition function 116. However, this embodiment is not limited to this. For example, the effect evaluation value may be calculated by a computer different from the medical information processing device 1, and the processing circuitry 11 may acquire the effect evaluation value from that computer. Alternatively, the processing circuitry 11 may acquire the effect evaluation value from a database that stores and manages effect evaluation values.

[0085] In some of the above embodiments, the processing circuitry 11 calculates the fairness index by implementing the fairness index acquisition function 117. However, this embodiment is not limited to this. For example, the fairness index may be calculated by a computer different from the medical information processing device 1, and the processing circuitry 11 may acquire the fairness index from the computer. Alternatively, the processing circuitry 11 may acquire the fairness index from a database that stores and manages fairness indexes.

[0086] If the effect depends not only on clinical judgment but also on features such as the subject's attributes or condition, it is possible to calculate the effect evaluation value for each feature separately, or to calculate the effect evaluation value using a model formula that expresses the relationship between the feature and the effect. This makes it possible to apply it to contextual bandits and sequential decision-making. For example, by using methods such as PSRL, BDQN, and SAC (Soft Actor Critic), it is possible to calculate the subjective optimal arm probability even if it is an estimated value of the effect.

[0087] (Summary) According to some of the above examples, the medical information processing device 1 according to this embodiment includes a processing circuit 11. The processing circuit 11 acquires effect evaluation values ​​for a plurality of medical decisions, acquires fairness index values ​​for the plurality of medical decisions, and determines a medical decision to be assigned to a target subject based on the effect evaluation values ​​and the fairness index values.

[0088] According to the above configuration, it is possible to determine the treatment decision for the allocation target using not only the effect evaluation value but also the fairness index value, which allows the selection of exploratory actions for multiple subjects in a fair manner.

[0089] According to at least one of the embodiments described above, it is possible to reduce unfairness among subjects in the allocation of clinical decisions while maintaining the magnitude of benefit to all subjects.

[0090] The term "processor" used in the above description refers to a circuit such as a CPU, a GPU, an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)). A processor realizes its functions by reading and executing a program stored in a memory circuit. Note that instead of storing a program in a memory circuit, a program may be directly embedded in the processor circuit. In this case, the processor realizes its functions by reading and executing the program embedded in the circuit. Alternatively, instead of executing a program, a function corresponding to the program may be realized by combining logic circuits. Note that each processor in this embodiment is not limited to being configured as a single circuit, but may be configured as a single processor by combining multiple independent circuits to realize its function. Furthermore, multiple components in FIG. 1 may be integrated into a single processor to realize its function.

[0091] Although several embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, modifications, and combinations of embodiments can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0092] 1 Medical information processing device 11 Processing circuit 12 Storage device 13 Input Devices 14. Communications equipment 15 Display equipment 111 Allocation function 112 Observation Function 113 Accumulation Function 114 Update function 115 Display control function 116 Effect evaluation value acquisition function 117 Fairness index value acquisition function 118 Allocation decision function

Claims

1. a first acquisition unit that acquires effect evaluation values ​​related to a plurality of medical decisions; a second acquisition unit that acquires fairness index values ​​related to the plurality of medical decisions; a decision unit that decides a medical decision to be assigned to a subject based on the efficacy evaluation value and the fairness index value; A medical information processing device comprising:

2. the fairness index value is a value based on the effectiveness evaluation value, the second acquisition unit calculates the fairness index value based on the effectiveness evaluation value; The medical information processing device according to claim 1.

3. the fairness index value is a value based on the effect evaluation value and a target value for the effect evaluation value set for each subject, the second acquisition unit calculates the fairness index value based on the effectiveness evaluation value and the target value; The medical information processing device according to claim 2.

4. The medical information processing device according to claim 3 , wherein the second acquisition unit sets the target value based on an effect evaluation value relating to a past medical treatment decision of the subject.

5. The medical information processing device according to claim 3 , wherein the second acquisition unit sets the target value based on an effect evaluation value regarding medical decisions selected by subjects other than the target subject.

6. The medical information processing device according to claim 1 , wherein the determination unit determines allocation candidates from among the plurality of medical decisions based on the effect evaluation value of each of the plurality of medical decisions, and determines the medical decision to be allocated based on the fairness index value of the allocation candidate.

7. The medical image processing apparatus according to claim 6 , wherein the determination unit determines the medical decision to allocate the allocation candidate when a fairness index value related to the allocation candidate satisfies a predetermined condition.

8. 2. The medical information processing device according to claim 1, wherein the determination unit stochastically samples a plurality of allocation candidates from the plurality of medical decisions according to a predetermined strategy based on a plurality of random numbers, and determines the medical decision to be allocated based on the fairness index values ​​of the plurality of allocation candidates.

9. The medical image processing apparatus according to claim 8 , wherein the determination unit determines the medical judgment to be allocated based on a comparison of fairness index values ​​between the allocation candidates.

10. A processor comprising: Obtaining effectiveness evaluation values ​​for multiple medical decisions, obtaining a fairness index value for the plurality of clinical decisions; determining a medical treatment decision to be assigned to the subject based on the efficacy evaluation value and the fairness index value; A medical information processing method comprising the steps of:

11. On the computer, A function to obtain effectiveness evaluation values ​​for multiple medical decisions; a function of acquiring a fairness index value for the plurality of medical decisions; a function of determining a medical decision to be assigned to a subject based on the efficacy evaluation value and the fairness index value; A medical information processing program that makes this possible.

Images