Result matching with personalized comparisons

Abstract

The dynamic user interface is configured to generate two types of information, which may be generated by complex computer systems such as expert systems and / or artificial intelligence. These types include "hypothesis" predictions that account for possible outcomes from alternative measures, and "similarity" analyses that display to the user how adjustments to the input data may affect the resulting predictions. The "hypothesis" prediction provides guidance for the user as to which measures may be most beneficial. 'Similarity' analysis helps the user to understand the consequence of the medical decision.

Description

Cross - Reference to Related Applications

[0001] This application claims priority and benefit of U.S. Provisional Patent Application No. 63 / 398,815, filed on August 17, 2022. This application is related to U.S. Provisional Patent Application No. 63 / 234,555, filed on August 18, 2021, and PCT Application No. PCT / US22 / 40782, filed on August 18, 2022. The disclosures of all of the above patent applications are hereby incorporated herein by reference. Background Art

[0002] Field of the Invention

[0003] Embodiments of the present invention are in the field of data presentation and, in some cases, in the field of customized presentation of medical outcome predictions (such as IVF).

[0004] Related Art

[0005] The black - box nature of some artificial intelligence systems can impede users' understanding and confidence in the generated results. Summary of the Invention

[0006] A dynamic user interface is configured to present to a user both a prediction of the expected outcomes that may occur from alternative measures and a comparison with a set of similar input data. The prediction is based on a comparison with a statistically relevant reference data set and can be made using statistical models and / or trained neural networks. The comparison with a set of similar input data is optionally configured to illustrate how changes in the input data can affect the probabilities of different outcomes. In the medical field, the comparison with a set of similar input data can include a comparison with patients having similar conditions (e.g., similar medical profiles).

[0007] Accordingly, the dynamic user interface provides two types of information that are generally difficult to express in an information - rich manner from the output of complex computer systems. These types include "hypothetical" predictions that illustrate possible outcomes from alternative measures and "similarity" analysis that shows users how adjustments to the input data may affect the final prediction. When performing similarity analysis between two people with similar profiles, this analysis can also be used to personalize the presented information because the user can be associated with people having similar profiles. This unique interface can be used to guide medical and other types of decisions made and increase users' understanding of the output of computer systems.

[0008] For illustrative purposes, a dynamic user interface is discussed in the context of a healthcare information system, particularly a system for supporting in vitro fertilization (IVF). The healthcare information system can include logic elements and / or other hardware configured to generate, manage, and / or deliver the presented information. For example, the healthcare information system can include storage, interactive user interface logic, cohort identification logic, profile comparison logic, outcome prediction logic, and / or other logic discussed herein.

[0009] As noted elsewhere herein, decision support can be applied in other healthcare applications and / or other types of non-healthcare applications.

[0010] Various embodiments of the present invention include: a data presentation system configured to present healthcare information to a patient, the system including: storage including non-transitory memory configured to store patient medical histories and outcomes for a plurality of past patients; outcome prediction logic configured to predict a healthcare outcome for a patient based on one or more healthcare measures; profile comparison logic configured to compare a past patient medical history with the patient's medical history; cohort identification logic configured to identify a cohort of past patients having a medical history similar to the patient's medical history; user interface logic configured to generate a user interface for presentation to the patient, the user interface including a representation of a prediction of a healthcare outcome generated by the outcome prediction logic and further including an interactive comparison between the patient's medical history and the medical histories of the cohort identified by the cohort identification logic; and a microprocessor configured to execute at least the user interface logic or the outcome prediction logic. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Illustrates a healthcare information system in accordance with various embodiments of the present invention.

[0012] Figures 2 to 3 Illustrates an example of a dynamic user interface that can be generated using the Figure 1 system therein.

[0013] Figures 4A to 4B Illustrates Figures 2 to 3 an expanded view of the dynamic user interface. DETAILED DESCRIPTION

[0014] Figure 1Illustrates a healthcare information system 100 according to various embodiments of the present invention. The healthcare information system 100 can be configured to present healthcare information to, for example, patients and / or healthcare providers. In various embodiments, the healthcare information system 100 includes a data presentation system 110, a network 115, and one or more client devices 120 (individually labeled as client device 120A, client device 120B, etc.). The presentation system 110 can be embodied on one or more computing devices, for example, on multiple network-based servers. The network 115 can include the Internet, a private network, a local network, and / or the like. The network 115 optionally includes wireless and / or wired communication channels. The client devices 120 can include computing devices of patients, healthcare providers, administrative departments, and / or third parties. For example, the client device 120 can include a patient's smartphone, a doctor's desktop computer, and / or the like. In some embodiments, the presentation system 110 is configured to communicate with the API of a medical record system via the network 115.

[0015] The presentation system 110 includes a storage 145, and the storage 145 includes non-transitory memory. The storage 145 can include any type of computing memory, including magnetic, optical, and / or electronic memory. The storage 145 is configured to store patient medical histories and outcomes for multiple past patients. The storage 145 can also be configured to store any combination of the logics discussed herein. For example, various logical elements of the presentation system 110 or the client device 120 as taught herein.

[0016] The patient medical histories stored in the storage 145 optionally include: the medical histories of the intended birth mother, egg donor, sperm donor, and / or embryos generated by fertilizing eggs from an egg donor. These medical histories can include any combination of medical data, and the medical data includes outcomes, decisions made, treatments, medications administered, patient characteristics (e.g., age, weight, location, health history, race, genetic history, etc.). The medical histories can include data about the current patient and / or past patients. For example, in some embodiments, the medical histories of past patients are stored elsewhere and used to train a machine learning system (e.g., the outcome prediction logic 125), while the data about the current patient is stored in the storage 145. The medical histories can further include: reproductive history, hormone levels, birth mother age, egg donor age, embryo growth rate, and / or embryo cleavage time. The medical histories can include any medical information taught in U.S. Provisional Patent Application No. 63 / 234555.

[0017] In various embodiments, the storage 145 is configured to store biomedical graphs, such as the biomedical graphs taught in U.S. Provisional Patent Application No. 63 / 234555, filed on August 18, 2021, and the outcome prediction logic 125 is configured to use the biomedical graphs to generate predictions, as taught therein.

[0018] The presentation system 110 further includes an outcome prediction logic 125. The outcome prediction logic 125 is configured to predict healthcare outcomes for a (current) patient. Specifically, the outcome prediction logic 125 is configured to predict in vitro fertilization (IVF) outcomes. The prediction is based on one or more healthcare measures. For example, the outcome prediction logic 125 may be configured to predict the outcome of a particular patient if the patient undergoes a single embryo transfer and to predict the outcome of the particular patient if the patient undergoes a double embryo transfer. In various embodiments, the outcome prediction logic 125 is configured to predict IVF outcomes based on the selection between alternative embryos, the selection of hormone therapy, the timing of embryo cell division, and / or the morphology of the embryo. In various embodiments, the outcome prediction logic 125 includes any system claimed in U.S. Provisional Patent Application No. 63 / 234555, filed on August 18, 2021.

[0019] Generally, the outcome prediction logic 125 is configured to provide comparable (quantitative) outputs. Specifically, the outcome prediction logic 125 is configured to make comparable predictions based on alternative healthcare measures (e.g., treatments). For example, the outcome prediction logic 125 may be configured to predict healthcare outcomes based on alternative medical treatments and / or provide a comparison between the expected outcomes of alternative medical treatments. In a specific example, the outcome prediction logic 125 may be configured to provide a quantitative prediction for a sequential embryo transfer and a quantitative prediction for a parallel (multiple embryo) transfer. These predictions are quantitative in that they can be numerically compared to each other. For example, one prediction gives a 25% probability of outcome "A", while another prediction gives a 45% probability of outcome "A".

[0020] In various embodiments, the outcome prediction logic 125 is configured to predict outcomes based on multiple healthcare measures and generate predictions based on various combinations of these healthcare measures. For example, a set of predictions may be generated based on healthcare measures related to: the number of eggs retrieved, the type of embryo transfer, and hormone supplementation. And the set may include multiple predictions, each based on a different combination of these measures. For example, a prediction for retrieving 4 eggs plus a sequential embryo transfer and not providing a specific hormone, a prediction for retrieving 8 eggs, multiple embryo transfers, and providing a specific hormone, and a prediction for retrieving 8 eggs plus a sequential embryo transfer and providing a specific hormone at a specific time during the IVF cycle.

[0021] The presentation system 110 also includes profile comparison logic 135. The profile comparison logic 135 is configured to compare a past patient medical history with the medical history of the current patient and identify groups. Such comparison can include medical history (events) and patient characteristics (e.g., age, parity, etc.). In various embodiments, the profile comparison logic 135 is configured to 1) compare medical profiles along multiple dimensions, 2) generate a (optionally weighted) distance between the medical profiles (cosine distance, Euclidean distance, or any other type of distance between vectors), 3) generate a value representative of the profile distance based on the weights of the profile features, and / or any combination thereof. For example, the profile comparison logic 135 can be configured to generate a cosine distance between medical profiles based on the weights of different patient characteristics and / or medical events. In various embodiments, the profile comparison logic 135 is configured to provide a list of the main differences between profiles, select / screen within a specific range (age), measure manageable differences and unmanageable differences in different ways, and / or the like. Manageable differences are differences that can be changed, such as a patient's weight or diet, while unmanageable differences are differences that cannot be changed, such as a patient's race or genetic history. In a specific example, the profile comparison logic 135 is optionally configured to make a distinction between static features (screened by age, race, past successful deliveries, past miscarriages, etc.) and dynamic features (hormone levels, weight, embryo hatching temperature, etc.) while making the comparison. In various embodiments, the profile comparison logic 135 can be used to compare the current patient with specific historical patients or aggregated historical patients.

[0022] The presentation system 110 optionally also includes group identification logic 130. The group identification logic 130 is configured to identify one or more groups of members of past patients with similar medical histories. The similarity can be weighted according to features related to the IFV results. This identification can be used to classify and / or group historical patients into groups. The group identification logic 130 can be configured to identify one or more groups of historical patients with medical histories and / or characteristics similar to those of the current patient.

[0023] Optionally, each group is distinguished by different medical measures taken in the historical patients. For example, the profile comparison logic 135 can be used to identify historical patients with medical profiles similar to those of the current patient (up to the point of fertility treatment). Then, the group identification logic 130 can be configured to identify the group of identified historical patients who have taken medical measure "B" and the group of identified historical patients who have taken medical measure "C".

[0024] Thus, the population identification logic 130 and the profile comparison logic 135 can be used in at least two ways. In a first method, the population identification logic 130 is used to divide historical patients into different populations, and the profile comparison logic 135 is used to identify which of these populations are similar to the current patient. In a second method, the profile comparison logic 135 is configured to identify historical patients similar to the current patient, and then the population identification logic 130 is used to divide these identified historical patients into different populations. For example, they can be divided into populations based on the medical treatment decisions made by the respective identified historical patients.

[0025] The population identification logic 130 optionally includes a trained machine learning system configured to divide patients into the populations discussed herein. The population identification logic 130 is optionally configured to perform a statistical analysis of patient data in order to divide patients into the populations discussed herein.

[0026] In an illustrative example, the profile comparison logic 135 and the population identification logic 130 can be used together as follows. The profile comparison logic 135 is first used to identify historical patients having a medical profile and characteristics similar to the current patient. Then, the population identification logic 130 is used to divide these identified historical patients into populations based on medical treatment decisions, e.g., what the historical patients did next in the stage of their fertility treatment, which is similar to the current stage of the current patient's fertility treatment. Each of these populations represents a different treatment course open to the current patient. The outcome prediction logic 125 can then be used to analyze each population to predict the outcome of a specific treatment decision.

[0027] A substantial amount of information provided by the outcome prediction logic 125, the profile comparison logic 135, and the population identification logic 130 is optionally communicated to the user (e.g., the current patient and the healthcare provider) using the user interface logic 140. The user interface logic 140 can include a variety of functions that allow the user to interactively explore the expected outcomes of alternative medical measures. Such presentation and exploration produce important technical results. Specifically, the user can obtain a better understanding of the consequences of specific medical measures; the complex relationship between the patient's profile, medical decisions / measures, and outcome probabilities is provided in a simplified and understandable manner; and the patient's confidence in the calculated probabilities can be increased. These results directly lead to an improvement in the medical decision-making process and allow the user to increase the probability of success of an IVF cycle, e.g., a successful live birth. The user interface(s) generated by the user interface logic 140 are generally configured to be displayed on the client device 120. For example, within a browser or a client application. The user interface is optionally based on HTML / XML and can communicate securely using https.

[0028] In various embodiments, the user interface generated by the user interface logic 140 includes a representation of a healthcare outcome prediction generated by the outcome prediction logic 125, controls configured to interactively compare the current patient's medical history with the medical histories of historical patients within a population, controls configured to display the consequences of specific medical decisions, and / or other information and controls. Figure 2 FIGS. illustrate examples of a dynamic user interface that can be generated using Figure 1 the system shown in (i.e., via the user interface logic 140 based on data generated by the outcome prediction logic 125, the profile comparison logic 135, and / or the population identification logic 130). The user interface logic 140 is optionally configured to generate a user interface for presentation to a patient, the user interface including a representation of a prediction of a healthcare outcome generated by the outcome prediction logic and further including an interactive comparison between the current patient's medical history and the medical histories of a single historical patient and / or a group of historical patients identified by the population identification logic 130.

[0029] Specifically referring to Figure 2 , a particular user interface 200 illustrates the "decision to consider" 210. This information refers to the specific medical decision being considered (the choice of a medical action). An illustrative example presents the treatment question "Should I consider sequential egg retrieval or proceed with embryo transfer?"

[0030] The user interface 200 also shows the percentage 215 of decisions made by a group of historical patients having a similarity to the current patient, e.g., decisions made within a group of historical patients similar to the current patient. Exemplary values indicate that 20% of historical patients stopped treatment, 30% of historical patients chose embryo transfer, and 40% of patients chose sequential egg retrieval.

[0031] The user interface 200 also includes a data match 220 between the current patient and a specific historical user or group thereof. This value indicates the degree of closeness of the specific historical patient or group thereof to the current patient. Optionally, the user interface logic 140 is configured such that clicking on the data match 220 produces a list of the greatest differences and / or similarities between the current patient and the specific historical patient or group thereof. The value included in the data match 220 is optionally generated by the profile comparison logic 135, as described elsewhere herein. In various embodiments, the user can select the weights of the data used in the comparison, filter the data being compared, and / or otherwise select how the value is calculated.

[0032] The user interface 200 also includes an add data option 225 configured to add information to the user's profile. In some embodiments, selecting this option will result in Figure 3 the interface shown in. This interface allows the user to upload medical files and / or edit their personal characteristics. In some embodiments, Figure 3The interface shown includes indications of user characteristics under the user's control, such as smoking habits or alcohol consumption that may improve IVF outcomes if changed. The user can vary these characteristics and examine how they affect IVF (or other medical) outcomes. In some embodiments, the user interface logic 140 is configured to suggest changes to the characteristics of the current patient under the current patient's control, and in particular those changes that are most likely to improve the outcome. In some embodiments, Figure 3 the interface shown is configured to suggest what additional data is desired to be added to the profile of the current patient. Medical records added to the patient's profile are optionally processed using a machine learning system configured to extract relevant information from the medical records. The uploaded records can include images, narrative healthcare provider notes, laboratory results, information from an electronic medical record system, and / or the like.

[0033] The user interface 200 also includes representatives 230 of more than two medical measures selected by historical patients. In Figure 2 the exemplary illustration, three measures are illustrated, represented by "User A", "User B", and "User C". These "users" can be actual historical users or representative aggregations (groups) thereof. Optionally, selecting one of the representatives 230 will result in further details about that representative, including, for example, the outcome probability of the medical measure, the individual historical patient profile, comments from the patient added to their profile, comments from the healthcare provider added to their profile, costs, and / or the like. In some embodiments, the user can scroll or use controls to view additional representatives 230. For example, by clicking on one of the percentages 215, the user interface logic 140 can be configured to display the representatives 230 of users who have taken a specific medical measure. Further, by selecting a specific representative 230, an illustrative patient profile 235 of the historical user (or their group) can be presented. These presented profiles can include highlighting the differences between the presented profile and the current user's profile. When presented to other users, all medical records and profiles are anonymized and do not include personal information.

[0034] The rendering system 110 optionally also includes anonymization logic 150. The anonymization logic 150 is configured to anonymize the medical history of past patients. The purpose of anonymization is to remove information that could be used to identify a specific patient.

[0035] The presentation system 110 optionally further includes training logic 155. The training logic 115 is configured to train a machine learning system that is incorporated in the outcome prediction logic 125, the population identification logic 130, and / or the profile comparison logic 135. For example, the training logic 115 can be configured to train a neural network using historical patient data (e.g., medical history including outcomes and / or patient characteristics). In some embodiments, the training logic 21 is configured to train the machine learning system based on data of a specific population of historical patients.

[0036] The presentation system 110 includes at least one microprocessor 190. The microprocessor 190 is configured to execute Figure 1 any one or more of the various logic elements shown therein. The microprocessor 190 can include, for example, a digital microprocessor configured to execute computing instructions. The microprocessor 190 can include electronic, optical, and / or quantum components. The presentation system 110 optionally includes multiple microprocessors 190 distributed across multiple computing devices.

[0037] Figures 4A to 4B Shows Figures 2 to 3 an expanded view of the dynamic user interface.

[0038] Certain embodiments are specifically shown and / or described herein. However, it should be understood that modifications and variations are covered by the above teachings and within the scope of the appended claims without departing from the spirit and intended scope. For example, the described systems and methods can be applied to other medical situations or decision-making processes. For example, the systems and methods can be applicable to medical situations involving: mental health, oncology / cancer, Alzheimer's disease, Parkinson's disease, autoimmune diseases, complex surgical evaluations for assessing risks and benefits, heart problems, when combined with wearable devices for early prediction of health problems or damage caused when combined with DNA, hyper-personalization of almost every treatment in terms of dosage and medications, and / or the like. The systems and methods described herein can also be applied to non-medical decision-making, including, for example, for career planning, college applications, investments, buying a house, choosing a school for a child, optimal nutrition or diet.

[0039] The embodiments discussed herein are illustrative embodiments of the present invention. Since these embodiments of the present invention are described with reference to the drawings, various modifications or adjustments to the described methods and / or specific structures can become apparent to those skilled in the art. All such modifications, adjustments, or variations that rely on the teachings of the present invention and that have improved the prior art through these teachings are considered to be within the spirit and scope of the present invention. Therefore, these descriptions and drawings should not be considered restrictive, as it should be understood that the present invention is in no way limited to only the illustrated embodiments.

[0040] The computing systems and / or logic mentioned herein can include integrated circuits, microprocessors, personal computers, servers, distributed computing systems, communication devices, network devices, or the like, as well as various combinations of the above. The computing system or logic can also include volatile and / or non-volatile memories, such as random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), magnetic media, optical media, nano media, hard disks, optical disks, digital versatile disks (DVDs), optical circuits, and / or other devices configured to store analog or digital information (such as in a database). Paper is expressly excluded from the computer-readable media used herein. The computer-implemented steps of the methods pointed out herein can include a set of instructions stored on a computer-readable medium, and when these instructions are executed, the computing system performs these steps. A computing system programmed according to instructions from program software to perform a specific function is a dedicated computing system for performing those specific functions. While performing those specific functions, the data operated on by the dedicated computing system is at least electronically saved in the buffer of the computing system, and with each change to the stored data, the dedicated computing system is physically changed from one state to the next state.

[0041] The "logic" discussed herein is expressly defined to include hardware, firmware, or software stored on a non-transitory computer-readable medium, or any combination thereof. The logic can be implemented in an electronic and / or digital device (e.g., a circuit) to produce a dedicated computing system. Any system discussed herein optionally includes a microprocessor, including electronic and / or optical circuits, configured to execute any combination of the logic discussed herein. The methods discussed herein optionally include the execution of the logic by the microprocessor.

Claims

1. A data presentation system configured to present healthcare information to a patient, the system comprising: Storage, including non-transitory memory, configured to store patient medical histories and outcomes for a plurality of past patients; Outcome prediction logic configured to predict a healthcare outcome for a (current) patient based on one or more healthcare measures; Profile comparison logic configured to compare the past patient medical histories with the medical history of the patient; Group identification logic configured to identify a member group of past patients having a medical history similar to the medical history of the patient; User interface logic configured to generate a user interface for presentation to the patient, the user interface including a representation of a prediction of a healthcare outcome generated by the outcome prediction logic and further including an interactive comparison between the patient's medical history and the medical histories of the groups identified by the group identification logic; And A microprocessor configured to execute at least the user interface logic or the outcome prediction logic.

2. The system according to claim 1, wherein, The outcome prediction logic is configured to predict in vitro fertilization (IVF) outcomes.

3. The system according to claim 2, wherein, The outcome prediction logic is configured to predict in vitro fertilization outcomes based on a selection between alternative embryos, a selection based on hormone therapy, a time of embryonic cell division, and / or a morphology of the embryo.

4. The system according to any one of the preceding claims, wherein, The outcome prediction logic includes any system claimed in U.S. Provisional Patent Application No. 63 / 234555, filed Aug. 18, 2021.

5. The system according to any one of the preceding claims, wherein, The outcome prediction logic is configured to predict healthcare outcomes based on alternative medical treatments and / or provide a comparison between the predicted outcomes of the alternative medical treatments.

6. The system according to any one of the preceding claims, wherein, The user interface is configured to compare the profile of the current patient with the profiles of historical patients and / or their groups.

7. The system according to any one of the preceding claims, wherein, The user interface is configured to compare the profile of the current patient with the profile of a similar historical patient therein, as determined by the profile comparison logic.

8. The system according to any one of the preceding claims, wherein, The user interface is configured to display the percentage of historical patients who selected different healthcare measures at a specific point in IVF treatment.

9. The system according to any one of the preceding claims, wherein The user interface is configured to display the similarity between the current patient and historical patients or their groups.

10. The system according to any one of the preceding claims, wherein, The user interface is configured to 11. The system according to any one of the preceding claims, wherein, The user interface includes a decision diagram and / or similar patient profiles and is configured for the current patient to navigate the diagram to explore predicted outcomes following different healthcare measures.

12. The system according to any one of the preceding claims, wherein, The profile comparison logic is configured for the user to select between comparing their profile with a group of historical users or comparing their profile with an individual historical user.

13. The system according to any one of the preceding claims, wherein, The profile comparison logic is configured to 1) compare medical profiles in multiple dimensions, 2) generate a (optionally weighted) cosine distance between medical profiles, and 3) generate a value representative of the profile distance based on the weights of profile features.

14. The system according to any one of the preceding claims, wherein, The profile comparison logic is configured to distinguish between static features (screened by age, race, previous fertility) and dynamic features (hormone levels, weight, embryo hatching temperature).

15. The system according to any one of the preceding claims, wherein, The group identification logic is configured to identify a member group of past patients having a medical history similar to the medical history of the patient, the similarity being weighted according to features related to IFV outcomes.

16. The system according to any one of the preceding claims, wherein, The patient medical history includes: the medical history of the intended birth mother, egg donor, and / or embryo generated by fertilization of an egg from the egg donor.

17. The system according to any one of the preceding claims, wherein, The patient medical history includes: reproductive history, hormone levels, birth mother age, egg donor age, embryo growth rate, and / or embryo cleavage time.

18. The system according to any one of the preceding claims, wherein, The storage is further configured to store the figures taught in U.S. Provisional Patent Application No. 63 / 234555, filed Aug. 18, 2021, and the result prediction logic is configured to use the figures to generate predictions as taught therein.

19. The system according to any of the preceding claims, further comprising anonymization logic configured to anonymize the past patient medical history.

Images