Methods for family planning using predictive models

Abstract

Described herein are systems and methods for implementing predictive models for optimizing fertility treatment for a patient. A method may include, via a first model having received first patient data, predicting a probability distribution for an egg outcome of a retrieval cycle for the patient. The method may then include, via a second model having received second patient data, predicting a success rate for a first fertility outcome for the patient. Next, the method may include providing, via a user interface, an indicator of a probability of that a second, different fertility outcome will result from the retrieval cycle based on the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome. The second fertility outcome may include a quantity of live births, such as at least two live births.

Description

ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 METHODS FOR FAMILY PLANNING USING PREDICTIVE MODELS CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Patent Application Serial No. 63 / 688,261, filed August 28, 2024, the content of which is incorporated herein by reference in its entirety for all purposes. TECHNICAL FIELD

[0002] This invention relates generally to the field of optimizing a fertility treatment for a patient based on their family goals by using models to inform a treatment plan for the patient. BACKGROUND

[0003] In vitro fertilization (IVF) and intrauterine insemination (IUI) are two of the most common assisted reproductive technologies (ART) used to help individuals and couples facing infertility. IVF involves the extraction of eggs from the ovaries, which are then fertilized with sperm in a laboratory setting, while IUI involves directly inserting sperm into a woman's uterus during ovulation to increase the likelihood of fertilization. Both methods have been instrumental in helping countless patients achieve their dream of parenthood. However, the complex nature of these treatments necessitates thorough fertility counseling to set realistic expectations for prospective patients.

[0004] A critical aspect of fertility counseling is estimating fertility outcomes for patients, such as a number of live births that a patients can expect from a given treatment cycle. Many patients seek out ART treatments in order to have multiple children. Traditionally, the fertility outcome estimates are conveyed as cumulative live birth rate (CLBR) predictions. CLBR refers to the likelihood of achieving a live birth after one or more cycles of ART, typically focusing on the probability of a first live birth. While CLBR can be a valuable metric, it does not encompass the broader goals many patients may have, such as the desire for more than one child. Indeed, many tools that predict CLBR are not configured to provide individualized predictions of any kind and may not even provide accurate CLBR predictions. These limitations, and many more, highlight the need for new methods of predicting fertility outcomes that align with patients' individual family planning goals. An approach that provides patients with tailored predictions on the likelihood of achieving their desiredATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 number of children would bolster fertility counseling by facilitating patient satisfaction, leading to better-informed decisions and more personalized fertility care. SUMMARY

[0005] Described herein are methods and systems for optimizing fertility treatment for a patient.

[0006] A method for optimizing a fertility treatment for a patient may include predicting a probability distribution for an egg outcome of a retrieval cycle for the patient based on a first model having received first patient data, predicting a success rate for a first fertility outcome for the patient based on a second model having received second patient data, and providing, via a report, a probability that a second, different fertility outcome will result from the retrieval cycle based on the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome.

[0007] In some variations, one or both of the first and second patient data may include one or more of: age, BMI, infertility diagnosis, number of prior full-term births, number of prior retrieval cycles, number of embryos transfers from retrieval cycle, and preimplantation genetic test (PGT) status of one or more embryos. In some variations, the first model may be trained on a first training data set and the second model is trained on a second training data set. The first training data set may include at least 1,000 prior-patient retrieval cycles, and the second training data set may include at least 10,000 different prior-patient retrieval cycles. In some variations, one or both of the first and second models may include a regression model. In some variations, the egg outcome may include a number of eggs, a number of oocytes, a number of mature eggs, a number of post-mature eggs, a number of fertilized eggs, a number of zygotes, a number of embryos, a number of blastocysts, a number of usable blastocysts, a number of euploid blastocysts, or a number of usable euploid blastocysts resulting from the retrieval cycle. In some variations, the first fertility outcome may be an embryo transfer. In some variations, the second fertility outcome may be a number of live births. The second fertility outcome may be at least two live births. In some variations, providing the probability that the second fertility outcome will result from the retrieval cycle may include combining the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome. In some variations, the first model may be configured to predict an egg outcome attrition rate for the patient based on the first patient data, and the predicted probabilityATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 distribution for the egg outcome may be further based on the predicted egg outcome attrition rate. The method may further include providing, via the report, an indicator of the predicted probability distribution for the egg outcome. In some variations, the probability distribution for the egg outcome may be a first probability distribution for a first egg outcome, and the probability that the second fertility outcome will result from the retrieval cycle may be a first probability that the second fertility outcome will result from a first retrieval cycle, and the method may further include predicting a second probability distribution for a second egg outcome of a retrieval cycle for the patient based on the first model, and providing, via the report, a second probability that the second fertility outcome will result from a second retrieval cycle for the patient based on the second predicted probability distribution for the second egg outcome and the predicted success rate for the first fertility outcome. Predicting the second probability distribution for the second egg outcome may include predicting an attrition rate for the patient based on third patient data, and combining the predicted attrition rate with a predicted range for the second egg outcome. The predicted attrition rate may be a rate of anti-mullerian hormone (AMH) decline, and the third data may include a future age of the patient. In some variations, the egg outcome may be a first predicted egg outcome, and the method may further include predicting a second egg outcome based on a third model having received third patient data, and predicting the first egg outcome based on the second predicted egg outcome. Predicting the second egg outcome may include determining a retention factor for the second predicted egg outcome and adjusting the second predicted egg outcome based on the retention factor. The second predicted egg outcome may include a number of eggs, a number of oocytes, a number of mature eggs, a number of fertilized eggs, or a number of zygotes, and the first predicted egg outcome comprises a number of embryos, a number of blastocysts, a number of usable blastocysts, a number of euploid blastocysts, or a number of usable euploid blastocysts resulting from the second predicted egg outcome. The method may further include receiving, via a user, the second fertility outcome, wherein the second fertility outcome comprises a desired fertility outcome for the patient. The method may further include determining a fertility treatment plan for the patient based on the predicted probability that the second fertility outcome will result from the retrieval cycle. In some variations, determining the fertility treatment plan may include one or more of: determining a type of fertility treatment for the patient, determining or adjusting a number of treatment cycles for the patient, determining or adjusting a start date for a treatment cycle for theATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 patient, determining an egg freezing protocol for the patient, determining an embryo freezing protocol for the patient, determining a PGT protocol for the patient, and determining an egg type for the patient. In some variations, determining the type of fertility treatment may include electing an in vitro fertilization (IVF) process or an intrauterine insemination (IUI) process. In some variations, determining or adjusting the number of treatment cycles may include increasing or decreasing a number of treatment cycles. In some variations, determining or adjusting the start date for the treatment cycle may include advancing or delaying the start date. In some variations, determining the egg freezing protocol may include electing to freeze or to not freeze one or more eggs of the patient. In some variations, determining the embryo freezing protocol may include electing to freeze or to not freeze one or more embryos of the patient. In some variations, determining the PGT protocol may include electing to test or to not test one or more embryos of the patient. In some variations, determining the egg type may include electing to use eggs of the patient or eggs of a donor during fertility treatment.

[0008] In some variations, the method may be performed by a non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform the method. In some variations, the report may be provided via a graphical user interface (GUI) configured to receive at least a portion of one or both of the first or second patient data as adjustable input parameters.

[0009] Another method for optimizing a fertility treatment for a patient may include predicting a probability for a fertility outcome of a retrieval cycle for the patient based on one or more models having received patient data, and providing, via a report, an indicator of the predicted probability for the fertility outcome. The patient data may include at least an infertility diagnosis and a transfer number from the retrieval cycle, and the fertility outcome may include one or more of plurality of live births and a plurality of pregnancies resulting from the retrieval cycle.

[0010] Another method for optimizing a fertility treatment for a patient may include predicting a first egg outcome of a retrieval cycle for the patient based on a first model having received first patient data, predicting an egg outcome attrition rate for the patient based on a second model having received second patient data, determining a probability distribution for a second egg outcome of the retrieval cycle based on the first predicted egg outcome and the predicted egg outcome attrition rate, where the second egg outcome develops from the first egg outcome, predicting a success rate for aATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 first fertility outcome for the patient based on a third model having received third patient data, predicting a probability for a second, different fertility outcome of the retrieval cycle for the patient based on the probability distribution for the second egg outcome and the success rate for the first fertility outcome, and providing, via a user interface, an indicator of the probability for the second fertility outcome of the retrieval cycle for the patient.

[0011] In some variations, determining the predicted probability distribution of number of embryos may include combining the predicted egg outcome and the predicted embryo attrition rate. In some variations, the eggs may include oocytes or zygotes (2PNs). In some variations, the embryos may include one or both of euploid blastocysts and usable blastocysts.

[0012] Another method for optimizing a fertility treatment for a patient may include receiving patient data and a first predicted number of eggs resulting from a retrieval cycle for the patient, predicting, via one or more models, an embryo attrition rate for the patient based on the patient data, determining, via the one or more models, a probability distribution of number of embryos resulting from the retrieval cycle based on the predicted number of eggs and the predicted embryo attrition rate, and providing, via a user interface, a probability that at least two live births will result from the retrieval cycle based on the predicted probability distribution of number of embryos.

[0013] In some variations, determining the predicted probability distribution of number of embryos may include combining the predicted egg outcome and the predicted embryo attrition rate. In some variations, the eggs may be oocytes or zygotes (2PNs). In some variations, the embryos may be one or both of euploid blastocysts and usable blastocysts.

[0014] Another method for optimizing a fertility treatment for a patient may include receiving patient data and a probability distribution of a number of embryos resulting from a retrieval cycle for the patient, predicting, via one or more models, an embryo transfer success rate for the patient based on the patient data, predicting, via the one or more models, a probability that at least two live births will result from the retrieval cycle for the patient based on the predicted probability distribution of number of embryos and the predicted embryo transfer success rate, and providing, via a user interface, an indicator of the probability that at least two live births will result from the retrieval cycle for the patient.

[0015] In some variations, determining the probability that at least two live births will result from the retrieval cycle may include combining the predicted probability distribution of number ofATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 embryos and the predicted embryo transfer success rate. In some variations, the embryos may include one or both of euploid blastocysts and usable blastocysts. In some variations, the predicted embryo transfer success rate may be a per-transfer euploid blastocyst success rate.

[0016] Yet another method for optimizing a fertility treatment for a patient may include predicting a success rate of a fertility outcome of a fertility treatment cycle for the patient based on one or more models having received patient data, where the patient data may include one or more of age, infertility diagnosis, and sperm parameters, and providing, via a user interface, an indicator of the success rate of the outcome of the fertility treatment.

[0017] In some variations, the sperm parameters may include one or more of sperm concentration and sperm motility. In some variations, the fertility outcome may include one or both of pregnancy and live birth. In some variations, the fertility treatment comprises intrauterine insemination (IUI).

[0018] Described herein are also methods and systems for graphically representing family planning predictions.

[0019] A method for graphically representing family planning predictions for a patient, may include: receiving, at a processor, a desired fertility outcome for the patient and patient data including one or more of: age, BMI, AMH level, infertility diagnosis, number of prior births, number of prior retrieval cycles, number of prior eggs frozen, and number of prior losses; determining a probability that the desired fertility outcome will result from each of one or more retrieval cycles for the patient based on the patient data and one or more models; generating a family planning report for the patient based on the probability that that the desired fertility outcome will result from each of the one or more retrieval cycles; and providing the family planning report via a graphical user interface (GUI). The GUI may include a first indicator representing the probability that the desired fertility outcome will result from each of the one or more retrieval cycles, a second indicator comprising a plurality of steps of a fertility process, and a plurality of third indicators each representing a predicted egg outcome for one of the plurality of steps.

[0020] In some variations, one or more of the first, second, and third indicators may include one or more of text, an image, a symbol, and a graph. In some variations, the second indicator may include a timeline. In some variations, the predicted egg outcome may include a number of eggs retrieved, a number of eggs frozen, a number of eggs thawed, a number of eggs fertilized, a number of blastocysts developed, or a number of euploids developed. In some variations, the processor mayATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 be configured to determine the probability that the desired fertility outcome will result from each of the one or more retrieval cycles, and to generate the family planning report. In some variations, the processor may be communicably coupled to a display configured to provide the GUI. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Non-limiting examples of various aspects and variations of the invention are described herein and illustrated in the accompanying drawings.

[0022] FIG.1 depicts an illustrative variation of a system for executing a family planning tool to generate family planning predictions and reports.

[0023] FIG.2A depicts an illustrative first configuration of an egg outcome model of a family planning tool. FIG.2B depicts an illustrative second configuration of the egg outcome model of FIG.2A.

[0024] FIG.3 depicts an illustrative configuration of an egg outcome attrition model of a family planning tool.

[0025] FIG.4 shows a result of a first exemplary implementation of an egg outcome attrition model of a family planning tool and a second exemplary implementation of the egg outcome attrition model.

[0026] FIG.5 depicts an illustrative configuration of a first fertility outcome success model of a family planning tool.

[0027] FIG.6 shows a result and analysis of an exemplary implementation of a first fertility outcome success model of a family planning tool.

[0028] FIG.7 depicts an illustrative configuration of a fertility attrition model of a family planning tool.

[0029] FIG.8 shows a result of an exemplary implementation of a fertility attrition outcome failure model of a family planning tool.

[0030] FIG.9 depicts an illustrative configuration of a fertility attrition model of a family planning tool.

[0031] FIG.10 shows a result and analysis of an exemplary implementation of a fertility outcome failure model of a family planning tool.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0032] FIG.11 depicts an illustrative configuration of a second fertility outcome success model of a family planning tool.

[0033] FIG.12A depicts an exemplary first configuration of a graphical user interface (GUI) dashboard for providing a family planning report. FIG.12B shows an exemplary second configuration of a portion of the GUI dashboard of FIG.12A.

[0034] FIG.13A depicts an exemplary first configuration of another GUI dashboard for providing a family planning report. FIG.13B shows an exemplary second configuration of a portion of the GUI dashboard of FIG.13A.

[0035] FIG.14 depicts another exemplary GUI dashboard for providing a family planning report related to egg freezing with previous cycle history.

[0036] FIG.15 depicts an exemplary portion of a family planning report related to preimplantation genetic testing (PGT) and provided on a GUI dashboard.

[0037] FIG.16 depicts another exemplary portion of a family planning report related to intrauterine insemination (IUI) treatment and provided on a GUI dashboard.

[0038] FIG.17 depicts another exemplary portion of family planning report related to donor fertility outcomes and provided on a GUI dashboard.

[0039] FIG.18 depicts another exemplary portion of a family planning report related to optimal dose of ovarian stimulation and provided on a GUI dashboard.

[0040] FIG.19 depicts another exemplary portion of a family planning report related to fertility treatment planning and provided on a GUI dashboard.

[0041] FIG.20 depicts a flow diagram of an illustrative method for optimizing fertility treatment for a patient using a predictive family planning tool.

[0042] FIG.21 depicts a flow diagram of another illustrative method for optimizing fertility treatment for a patient using a series of models of a predictive family planning tool.

[0043] FIG.22 depicts a flow diagram of an illustrative method for graphically representing family planning predictions from a predictive family planning tool. a optimizing fertility treatment for a patient using a series of models of a predictive family planning tool.

[0044] FIG.23 depicts a flow diagram of another illustrative method for optimizing fertility treatment for a predictive family planning tool.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 DETAILED DESCRIPTION

[0045] In vitro fertilization (IVF) and intrauterine insemination (IUI) are among the most widely utilized assisted reproductive technologies (ART), helping countless individuals and couples overcome infertility challenges. IVF involves fertilizing eggs with sperm in a laboratory and transferring the resulting embryo(s) into the uterus, while IUI entails placing sperm directly into the uterus during ovulation to facilitate fertilization. Despite the success of these treatments, the inherent complexity of ART underscores the critical importance of comprehensive fertility counseling ^ especially during initial consultations ^ to educate patients on the fertility process and establish realistic expectations.

[0046] These consultations serve as a foundational step in the fertility treatment process by establishing clear expectations regarding likely outcomes and treatment steps. Providing accurate, individualized predictions during such consultations reduces uncertainty, improves patient adherence to treatment protocols, and facilitates timely initiation of assisted reproductive technologies (ART). Machine learning may be an important tool for guiding these consultations. For example, models have been used to predict cumulative live birth rate (CLBR) for a patient, which estimates the likelihood of achieving a first live birth after an ART cycle. However, CLBR predictions alone may not be sufficient for some patients, particularly those desiring multiple children. As described herein throughout, it may be beneficial to instead predict success rates for multiple fertility outcomes, such as multiple live births, pregnancies, or embryo transfers, resulting from a given treatment cycle, to offer a more complete picture of the patient^s prospects. These predictions (referred to herein as ^family planning predictions^) may additionally improve fertility treatment by aiding patients and medical professionals in decision making throughout treatment. For example, based on the family planning predictions, patients and / or their medical team may choose a particular treatment or combination of treatments (e.g., IVF with egg freezing, IVF without egg freezing, IUI, etc.), adjust timing for a treatment and or / one or more actions within that treatment (e.g., a start date for the treatment, earlier egg retrieval date, etc.), a number of treatment cycles to undergo (e.g., two, three, or more treatment cycles), and the like.

[0047] Further, traditional CLBR calculations process patient data through a single predictive pathway and thus lose information about intermediate outcomes. In contrast, the prediction process herein may maintain probability distributions through each developmental stage to preserveATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 uncertainty information. The predictions may be made using a cascading model that enables stage- specific optimization. That is, each model may be independently validated and updated based on the process it represents.

[0048] Moreover, traditional predictive tools often fail to provide contextual data and detailed explanations associated with the predictions, leaving patients with an incomplete understanding of their prognosis. Providing reports during consultations, such as family planning reports that include both the predicted success rates for desired fertility outcomes and contextual explanations, may enhance patient satisfaction. For example, a family planning report including predictions for outcomes of each step of a fertility process, as well as the predictions for success rates of an overall fertility outcome (e.g., a number of live births per treatment cycle), may empower patients with a more detailed breakdown of predicted treatment outcomes at each stage, thereby supporting realistic expectation setting. By incorporating stage-specific probability distributions into the reports, the disclosed systems allow medical professionals to tailor treatment plans that are optimized for the patient^s individual fertility profile and long-term family goals.

[0049] Further, when patients are well-informed and have realistic expectations, they may be more likely to make quicker, more confident decisions about their treatment. Accordingly, a medical establishment using such reports may initiate treatment sooner, ultimately benefiting both the patient and the practice. Furthermore, the family planning reports may facilitate more effective communication between patients and medical professionals. For instance, when a medical professional needs to discuss the lower likelihood of achieving a desired family size with an older patient or someone with a history of infertility, an objective report may standardize communication of outcome probabilities, reducing variability across different practitioners. This standardization may improve consistency of clinical counseling and enhance long-term care continuity. In particular, the objective presentation of outcomes may facilitate informed consent and improve compliance with recommended ART protocols. The report may serve as a neutral, informative tool also fosters trust and understanding between the patient and the medical professional, enhancing patient satisfaction and long term care.

[0050] Accordingly, providing accurate, tailored family planning predictions, such as via explanatory family planning reports that include fertility outcome predictions for a patient^s desired family goals, may not only improve patient understanding of treatment outcomes but also enableATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 more efficient ongoing management of multi-cycle fertility treatments, thereby reducing delays in initiating or modifying protocols.

[0051] Described herein are systems and methods that implement a predictive family planning tool to generate family planning predictions, such as a predicted probability of achieving a desired fertility outcome, and / or one or more predicted egg outcomes. A fertility outcome may refer to a successful outcome of a completed ART treatment cycle, such as live birth, pregnancy (e.g., a clinical pregnancy), or embryo transfer (e.g., successful embryo transfer). An egg outcome may result from a step within the treatment cycle, or a developmental stage of an egg or embryo. For example, nonlimiting examples of an egg outcome may include a number of eggs, a number of oocytes, a number of mature (MII) eggs, a number of post-mature eggs, a number of fertilized eggs, a number of frozen eggs, a number of thawed eggs (e.g., eggs that survive thawing), a number of zygotes, a number of embryos, a number of blastocysts, a number of usable blastocysts, a number of euploid blastocysts, or a number of usable euploid blastocysts.

[0052] In particular, a predictive family planning tool may include one or more models configured to be implemented together (e.g., a plurality of predictive models, which may be implemented as a series of models run sequentially) to make family planning predictions. Each of the one or more models may be trained with datasets including on the order of hundreds, thousands, tens of thousands, or hundreds of thousands of prior patient treatment cycles (e.g., retrieval and / or IUI cycles), resulting in accurate and personalized family planning predictions and thus patient satisfaction throughout the treatment process.

[0053] The family planning predictions may be used to generate a family planning report for a patient and / or medical professional, which may be verbal, written (e.g., on paper) and / or digital. For example, the family planning report may be provided in the form of a verbal report (e.g., via a medical professional), a printed report, and / or an electronic report (e.g., a PDF, an email and / or via a graphical user interface (GUI) provided via a display). The family planning report may anticipate and address common patient questions by providing comprehensive explanations about the family planning predictions provided and about various treatment options. For example, the family planning report may indicate a probability of success of the patient achieving a desired fertility outcome over one or more ART treatment cycles. The family planning report may additionally or alternatively show the patient^s predicted egg outcome following each of a plurality of steps of aATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 fertility treatment. For example, with respect to an IVF retrieval cycle, the family planning report may provide a predicted egg outcome following one or more of egg retrieval, egg freezing, egg thawing, euploid embryo selection, and embryo transfer. Predicted egg outcomes may also be shown for one or more stages of egg and / or embryo development, such as for mature egg development, zygote development, embryo development, and / or blastocyst development.

[0054] In some variations, the family planning predictions and optionally the family planning report may be used to determine a treatment plan for a patient and / or to perform one or more treatment steps. For example, a medical professional may use one or more family planning predictions to inform decisions related to one or more of: a type of ART treatment (IVF, IUI), a treatment protocol (e.g., number of IVF retrieval cycles, a number of egg freezing cycles, a number of IUI cycles, a number of embryo biopsies, a number of embryo transfers), use of preimplantation genetic testing (PGT), use of donor eggs, use of donor sperm, a medication dosage (e.g., of ovarian stimulation medication), a hormonal trigger day, and / or the like. The family planning prediction(s) and / or report may be used to adjust a planned fertility treatment for a patient, such as to adjust one or more of: a treatment timing (e.g., accelerate or delay a start date for a given treatment), a treatment protocol (e.g., increase or decrease a number of treatment cycles), a medication dosage (e.g., increase or decrease a starting dosage of ovarian stimulation medication), a hormonal trigger day timing (e.g., accelerate or delay the hormonal trigger day), and / or the like. As another example, based on the family planning prediction(s) and / or report, the patient may elect to use one or more of preimplantation genetic testing (PGT), donor eggs, and donor sperm to increase their chances of achieving their desired fertility outcomes. As will be described in detail herein, in some variations, a family planning report may provide indicators of a success rate for a patient^s desired fertility outcome given one or more of the following conditions: no PGT, use of PGT, no donor eggs, use of donor eggs, no donor sperm, and use of donor sperm. Accordingly, the patient may be able to compare the success rate of their desired fertility outcome under some or all of the aforementioned conditions, allowing them to make a more informed and quicker decision about enhancing their fertility treatment with PGT, donor eggs, and / or donor sperm.

[0055] Similarly, in some variations, the family planning predictions herein may be used to actively treat a patient. Treating the patient may include administering medication, such as ovarian stimulation medication or a trigger shot, and / or performing a procedure, such as egg retrieval,ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 embryo biopsy, embryo transfer, and / or the like in order to support an optimized fertility process for the patient. That is, in some variations, the family planning predictions and / or report may directly inform initiation of an ART-related surgical procedure and / or administration of one or more medications, such as such as one or more ovarian stimulation medications or a hormonal trigger shot. Put differently, the family planning predictions and / or report may provide motivation for or otherwise result in performance of an ART-related surgical procedure and / or administration of one or more medications to the patient. This direct link between a predictive output and medical intervention underscores the practical application of the disclosed system in optimizing ART outcomes.

[0056] Variations of systems and methods for family planning using the predictive family planning tool are described below. I. System

[0057] FIG.1 depicts an exemplary variation of a system 100 for providing family planning predictions. The system 100 may be configured to implement the predictive family planning tool herein to provide such predictions, such as in the form of a report for review by a patient and / or medical professional. The system 100 may access and / or retrieve data from reliable electronic medical records (EMR) 112, such as one or more patient datapoints, to implement the predictive family planning tool using the data retrieved from EMR 112.

[0058] The EMR 112 may be a database such as eIVF^ patient portal, Artisan^ fertility portal, Babysentry^ management system, EPIC^ patient portal, IDEAS^ from Mellowood Medical, etc., or any suitable electronic medical record management software. In some variations, the EMR 112 may be associated with a specific clinic. In such variations, the EMR 112 may be configured to interface with one or more servers associated with the specific clinic. In some variations, the EMR 112 may be hosted on a cloud-based platform (e.g., Microsoft Azure®, Amazon® web services, IBM® cloud computing, etc.).

[0059] In some variations, the EMR 112 from a specific clinic may not be shared with other hospitals, pharmacies, practitioners, etc. Additionally, or alternatively, the EMR 112 may be configured to access databases associated with each clinic. The EMR 112 may automatically extract relevant information from a patient^s chart and match it against a database of de-identified medicalATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 records (e.g., database 114 described below). Accordingly, the relevant data across several entities (e.g., clinics, hospitals, pharmacies, practitioner, etc.) may be extracted from the EMR 112 without compromising the privacy of the patients (e.g., by maintaining Health Insurance Portability and Accountability Act regulations).

[0060] The EMR 112 may be accessed via a computing device. Some non-limiting examples of the computing device include computers (e.g., desktops, personal computers, laptops etc.), tablets and e-readers (e.g., Apple iPad®, Samsung Galaxy® Tab, Microsoft Surface®, Amazon Kindle®, etc.), mobile devices and smart phones (e.g., Apple iPhone®, Samsung Galaxy®, Google Pixel®, etc.), etc. For example, EMR 112 may be stored on a memory associated with the computing device. Alternatively, the EMR 112 may be accessed online through a web browser (e.g., Google®, Mozilla®, Safari®, Internet Explorer®, etc.) rendered on the computing device. In yet another alternative variation, the EMR 112 may be stored on a third-party database that may be accessed via the computing device.

[0061] Patient data may be extracted from the EMR 112. The patient data extracted from the EMR 112 may refer to: (1) data associated with one or more patients that may include the description, content, values of records, a combination thereof, and / or the like; and / or (2) metadata providing context for the said data. For example, patient data extracted from the EMR 112 may include one or both the data and metadata associated with patient records.

[0062] The patient data may be associated with a primary patient and, in some variations, the patient^s partner, egg donor, and / or sperm donor. Some non-limiting examples of patient data extracted from the EMR 112 may include: age (e.g., current age, age at time of given retrieval cycle), body mass index (BMI), infertility diagnosis, prior fertility outcomes (e.g., number of prior full-term births, pregnancies, and / or embryo transfers), prior losses (e.g., number of prior miscarriages, surgical abortions, medically-induced abortions, and / or still-births), prior egg outcomes (e.g., number of mature (MII) eggs, number of frozen eggs, number of thawed eggs (e.g., eggs that survive thawing), number of embryos, number of blastocysts, number of euploid blastocysts, number of usable blastocysts / euploid blastocysts), treatment cycle number (i.e., retrieval cycle number or insemination cycle number based on prior and / or predicted number of retrieval cycles or IUI cycles, respectively), transfer number from retrieval cycle, storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh orATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 frozen), PGT status of one or more embryos (e.g., blastocysts or euploid blastocysts), race, ethnicity, surgical history (e.g., prior uterine surgery history), medications, measurements such as measurements of anti-mullerian hormone (AMH), estradiol (E2), luteinizing hormone (LH), progesterone (P4), follicle stimulating hormone (FSH), and / or antral follicle count (AFC), sperm parameters (e.g., concentration and / or motility), and / or the like. In some variations, an infertility diagnosis for a patient may account for one or more of: male factor, endometriosis, ovulation disorders, polycystic ovaries, diminished ovarian reserve, tubal factor, unexplained infertility, recurrent pregnancy loss, genetic diseases, and the like.

[0063] The EMR may be communicably coupled to a controller 102 configured to receive the patient data. In some variations, the controller 102 may include one or more servers and / or one or more processors running on a cloud platform (e.g., Microsoft Azure®, Amazon® web services, IBM® cloud computing, etc.). The server(s) and / or processor(s) may be any suitable processing device configured to run and / or execute a set of instructions or code, and may include one or more data processors, image processors, graphics processing units, digital signal processors, and / or central processing units. The server(s) and / or processor(s) may be, for example, a general purpose processor, a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), and / or the like.

[0064] Similarly, the controller 102 may be communicably coupled to a database 114 which may include a plurality of databases. In some variations, the database 114 may include a public database, such as the SART CORS database. Additionally, or alternatively, the database 114 may include a private database. The database 114 may store hundreds to thousands of prior patient records (e.g., de-identified records). Like the patient data, prior patient data may include, for a given prior patient, one or more of: age (e.g., current age, age at time of given retrieval cycle), BMI, infertility diagnosis, prior fertility outcomes (e.g., number of full-term births, pregnancies, embryo transfers), prior losses (e.g., number of prior miscarriages, surgical abortions, medically-induced abortions, and / or still-births), prior egg outcomes (e.g., number of mature (MII) eggs, number of frozen eggs, number of thawed eggs (e.g., eggs that survive thawing), number of embryos, number of blastocysts, number of euploid blastocysts, number of usable blastocysts / euploid blastocysts) treatment cycle number (i.e., retrieval cycle number or insemination cycle number based on prior and / or predicted number of retrieval cycles or IUI cycles, respectively), transfer number fromATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 retrieval cycle, PGT status of one or more embryos (e.g., blastocysts or euploid blastocysts), storage status of one or more embryos (e.g., fresh or frozen), race, ethnicity, surgical history (e.g., prior uterine surgery history), medications, and measurements such as measurements of AMH, E2, LH, P4, FSH, and / or AFC, sperm parameters (e.g., concentration and / or motility), and / or the like.

[0065] In some variations, information and family planning predictions (e.g., egg outcome predictions and / or fertility outcome success rate predictions) may be stored in the database 114. These predictions may be accessed from the database 114 to further improve the accuracy of the models herein.

[0066] In some variations, the controller 102 may include a processor 104 (e.g., CPU). The processor 104 may be any suitable processing device configured to run and / or execute a set of instructions or code, and may include one or more data processors, image processors, graphics processing units, physics processing units, digital signal processors, and / or central processing units. The processor 104 may be configured to run and / or execute application processes and / or other modules, processes and / or functions associated with the system and / or a network associated therewith. For example, the processor 104 may be configured to run one or more family planning models to generate one or more predictions related to family planning (e.g., egg outcome and / or fertility outcome predictions). The underlying device technologies may be provided in a variety of component types (e.g., MOSFET technologies like complementary metal-oxide semiconductor (CMOS), bipolar technologies like emitter-coupled logic (ECL), polymer technologies (e.g., silicon- conjugated polymer and metal-conjugated polymer-metal structures), mixed analog and digital, and / or the like. The processor 104 may be configured to implement the models and methods herein using software (executed on hardware), hardware, or a combination thereof. Hardware modules may include, for example, a general-purpose processor (or microprocessor or microcontroller), a field programmable gate array (FPGA), and / or an application specific integrated circuit (ASIC). Software modules (executed on hardware) may be expressed in a variety of software languages (e.g., computer code), including structured text, typescript, C, C++, C#, Java®, Python, Ruby, Visual Basic®, and / or other object-oriented, procedural, or other programming language and development tools. Examples of computer code include, but are not limited to, micro-code or micro-instructions, machine instructions, such as produced by a compiler, code used to produce a web service, and files containing higher-level instructions that are executed by a computer using an interpreter. AdditionalATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 examples of computer code include, but are not limited to, control signals, encrypted code, and compressed code.

[0067] In some variations, the processor 104 may be configured to transmit and / or receive data and / or other signals from a storage medium such as the EMR 112, memory 106, or database 114.

[0068] The controller 102 may additionally include a memory 106 configured to store data and / or information. In some variations, the memory 106 may be configured to store any received data and / or data generated by the processor 104, such as one or more predictions related to family planning (e.g., egg outcome and / or fertility outcome predictions). In some variations, the memory 106 may include one or more of a random-access memory (RAM), static RAM (SRAM), dynamic RAM (DRAM), a memory buffer, an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), a read-only memory (ROM), flash memory, volatile memory, non-volatile memory, combinations thereof, and the like. In some variations, the memory 106 may store instructions to cause the processor to execute modules, processes, and / or functions associated with the device, such as image processing, image display, sensor data, data and / or signal transmission, data and / or signal reception, and / or communication. Some embodiments described herein may relate to a computer storage product with a non-transitory computer-readable medium (also may be referred to as a non-transitory processor-readable medium) having instructions or computer code thereon for performing various computer-implemented operations. The computer-readable medium (or processor-readable medium) is non-transitory in the sense that it does not include transitory propagating signals per se (e.g., a propagating electromagnetic wave carrying information on a transmission medium such as space or a cable). The computer code (also may be referred to as code or algorithm) may be those designed and constructed for the specific purpose or purposes. The memory 106 may be configured to store data temporarily or permanently.

[0069] The controller 102 may additionally include a display 108. The display 108 may be configured to provide a graphical user interface (GUI) 110. In some variations, a display may include at least one of a light emitting diode (LED), liquid crystal display (LCD), electroluminescent display (ELD), plasma display panel (PDP), thin film transistor (TFT), organic light emitting diodes (OLED), electronic paper / e-ink display, laser display, and / or holographic display. In some variations, the GUI 110 may be configured to show a family planning report, such as a report including one or more family planning predictions (e.g., egg outcome and / or fertilityATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 outcome success predictions) and / or information related to family planning. Thus, one or more family planning predictions, patient data, prior patient data, and / or related information may be transmitted to the GUI 110. The GUI 110 may show such predictions, data, and / or information in an interpretable format for a patient and / or medical professional. For example, the GUI 110 may provide, via the display 108, a first indicator for each of one or more predicted fertility outcomes, and a second indicator for each of one or more predicted egg outcomes. The second indicator may further include a plurality of steps of a fertility process, such as, for example, a timeline showing consecutive steps of an IVF and / or IUI process. One or both of the first and second indicators indicator may include one or more of text, an image, a symbol, and a graph. In some variations, the GUI 110 may be configured to receive user input (e.g., via a patient and / or medical professional), such as patient data and / or a patient^s desired fertility outcome. Such inputs may be used by the processor 104 to generate one or more family planning predictions for the patient.

[0070] In some variations, the GUI 110 may improve the functionality of an electronic medical record system by providing dynamic, interactive decision support. For example, unlike traditional reports that merely display success rates, the GUI 110 may enable patients and medical professionals to adjust one or more inputs, such as desired number of children, use of preimplantation genetic testing, or donor gametes, and receive updated outcome probabilities in real time. This dynamic functionality may provide a technical improvement to the operation of fertility management software systems, as it enables immediate recalculation of treatment scenarios without requiring repeated manual entry of data into separate tools. Such improvements may allow medical professionals to reduce consultation time and increase accuracy in treatment planning as compared to conventional reports. Exemplary variations of GUIs will be described in more detail herein. 1. Predictive Family Planning Tool

[0071] The predictive family planning (PFP) tool may be used to predict whether a patient will successfully achieve their desired family goals (i.e., desired number of children) via an ART treatment cycle. That is, the predictions may be a probability that a particular fertility outcome, such as a live birth (e.g., at least two live births, at least three live births, at least four live births or more), will result from an IVF retrieval cycle or IUI cycle for the patient.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0072] The fertility outcome may refer to a successful outcome of a completed ART treatment, such as live birth, pregnancy (e.g., a clinical pregnancy), or embryo transfer. In some variations, the PFP tool may be configured to predict a success rate for a plurality (e.g., two, three, four or more) different fertility outcomes, such as for any combination of a first fertility outcome (e.g., embryo transfer), a second fertility outcome (e.g., pregnancy), and a third fertility outcome (e.g., live birth). The PFP tool may be configured to receive (e.g., via user input to a GUI) a desired type(s) and / or number(s) of fertility outcomes so that a patient^s specific family planning goals may be assessed. For example, an initial, user-adjustable (via, e.g., the GUI) input to the PFP tool may include a quantity of fertility outcomes (e.g., one live birth, two live births, three live births, four live births), which may be presented to a user as a desired number of children. Accordingly, the PFP tool may predict a success rate for a plurality of fertility outcomes for the patient, such as for two, three, four, five, or more than five fertility outcomes (e.g., two live births, three live births, four live births, five live births, or more than five live births). In some variations, the PFP tool may be implemented a plurality of times to make predictions for a corresponding plurality of treatment cycles for the patient. For example, the PFP tool may predict a success rate for a fertility outcome for the patient over two, three, four, five, or more than five retrieval cycles. Because each cycle after the first would occur farther in the future, making predictions for a plurality of retrieval cycles may help patients understand how the probability of success of their desired fertility outcome changes over time. In some variations, the PFP tool may be implemented a plurality of times to make predictions for a single treatment cycle. For example, using the same inputs for each execution of the PFP tool, the PFP tool may be configured to generate a single, enhanced prediction by averaging the individual predictions.

[0073] The PFP tool may comprise a cascaded model configured to propagate predictions through successive stages, where each model's output serves as input to subsequent models. In some variations, this architecture may at least in part mirror the biological progression of fertility treatment, where outcomes at each stage (egg retrieval, fertilization, embryo development) influence subsequent probabilities. This architecture may also enable mid-treatment recalculations. For example, if a patient has completed a step of a fertility treatment protocol (e.g., egg retrieval), the PFP tool may be configured to begin predictions from just after that step (e.g., from egg retrieval) using actual rather than predicted values (e.g., actual egg outcomes).ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0074] Additionally, the PFP tool may predict one or more egg outcomes for the patient, where an egg outcome may result from a step of the ART treatment process or a developmental stage of an egg or embryo (e.g., a number of oocytes that mature into MII eggs, a number of blastocysts that fully develop into embryos, etc.). For example, the egg outcome may include one or more of: a number of eggs, a number of oocytes, a number of mature (MII) eggs, a number of post-mature eggs, a number of fertilized eggs, a number of frozen eggs, a number of thawed eggs (e.g., eggs that survive thawing), a number of zygotes, a number of embryos, a number of blastocysts, a number of usable blastocysts, a number of euploid blastocysts, and a number of usable euploid blastocysts. A predicted egg outcome may be used as an input for one or more models of the PFP tool, and / or may be shown to a patient (via a GUI) to provide information about their potential fertility treatment and context for their predicted probabilities of fertility outcome success. In some variations, a range for an egg outcome may be predicted. For example, the PFP tool may be configured to predict a distribution for an egg outcome and generate a range for the predicted egg outcome based on, for example, the standard deviation of the distribution. Providing a range of the predicted egg outcome (via, for example, a family planning report) may better prepare a patient for variance in their actual egg and associated fertility outcomes.

[0075] Additionally, or alternatively, in some variations, the PFP tool may predict one or more fertility outcomes for a patient based on a prior egg outcome for the patient, such as a number of eggs and / or embryos currently stored (e.g., frozen) for the patient. These predictions may, in some variations, be compared to one or more predicted fertility outcomes for the patient that are based on one or more predicted, future egg outcomes (e.g., from one or more future retrieval cycles for the patient). That is, the fertility outcome prediction(s) based on the patient^s prior egg outcome(s) may be provided along with the fertility outcome prediction(s) based on the patient^s predicted egg outcome(s) so that the patient may understand how undergoing one or more additional treatment cycles may impact their chances of achieving their desired fertility outcome. In some variations, the fertility outcome prediction(s) based on the patient^s prior egg outcome(s) may be provided without fertility outcome prediction(s) based on predicted egg outcome(s).

[0076] In some variations, the PFP tool may be configured to provide treatment predictions for a patient, such as predicted optimal dose of ovarian stimulation medication for the patient and / or a predicted optimal hormonal trigger day for the patient. Additional systems, models, and methods forATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 providing optimal dose and trigger day predictions, and aspects thereof, may be provided in U.S. Pat. No.11,735,302, the content of each of which is hereby incorporated by reference herein in its entirety.

[0077] To make family planning predictions, the PFP tool may utilize one or a series of models, which are described in detail below. Each model may be configured to receive unique inputs, such as unique patient data and / or outputs from one or more of the other models, or they may be configured to receive one or more of the same inputs. For example, a first model of the PFP tool may be configured to receive first patient data, a second model of the PFP tool may be configured to receive second patient data, a third model of the PFP tool may be configured to receive third patient data, and so on. In some variations, any combination of the first, second, third, and any additional data may at least partially overlap (i.e., may include at least one of the same datapoints), may completely overlap (i.e., may include all of the same datapoints), or may be completely different (i.e., may include none of the same datapoints).

[0078] The patient data may include one or more of: age (e.g., current age, age at time of given retrieval cycle), BMI, infertility diagnosis, prior fertility outcomes (e.g., number of full-term births, pregnancies, embryo transfers), prior losses (e.g., number of prior miscarriages, surgical abortions, medically-induced abortions, and / or still-births), prior egg outcomes (e.g., number of mature (MII) eggs, number of frozen eggs, number of thawed eggs (e.g., eggs that survive thawing)), number of embryos, number of blastocysts, number of euploid blastocysts, number of usable blastocysts / euploid blastocysts) treatment cycle number (i.e., retrieval cycle number or insemination cycle number based on prior and / or predicted number of retrieval cycles or IUI cycles, respectively), transfer number from retrieval cycle, PGT status of one or more embryos (e.g., blastocysts or euploid blastocysts), storage status of one or more eggs and / or embryos (e.g., fresh or frozen), race, ethnicity, surgical history (e.g., prior uterine surgery history), medications, and measurements such as measurements of AMH, E2, LH, P4, FSH, and / or AFC, sperm parameters (e.g., concentration and / or motility), and / or the like. In some variations, for a patient who has undergone one or more prior retrieval cycles, the PFP tool (e.g., one or more models thereof) may be configured to receive the patient^s prior egg outcome(s) (e.g., number of eggs retrieved, number of eggs frozen, number of eggs that survive thawing, number of embryos frozen, number of embryos that survive thawing, etc.) to predict their chance of achieving another egg outcome and / orATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 fertility outcome at a future stage of their fertility treatment given their prior egg outcome(s). In some variations, the patient data may include data associated with the patient^s partner, egg donor, and / or sperm donor.

[0079] In some variations, the PFP tool may include one or more regression models (e.g., linear regression, Poisson regression, binomial regression, logistic regression, CatBoost regression, etc.). Additionally, or alternatively, the PFP tool may include a neural network (e.g., recurrent neural network, LSTM, etc.). In some variations, the PFP tool may be configured to capture complex and / or nonlinear relationships between input parameters using decision trees and / or boosting (e.g., gradient boosting to combine multiple predictive models). In some variations, the PFP tool may be configured to analyze and optimize for categorical features, missing values, and / or overfitting.

[0080] In some variations, a PFP tool may be trained using current and / or prior patient data. A prior patient may be any patient who is not the patient of interest (e.g., past and / or current patients who are not who are not the patient of interest). Like the patient data, the prior patient data may include one or more of: age (e.g., at time of retrieval cycle), BMI, infertility diagnosis, prior fertility outcomes (e.g., number of full-term births, pregnancies, embryo transfers), prior losses (e.g., number of prior miscarriages, surgical abortions, medically-induced abortions, and / or still-births), prior egg outcomes (e.g., number of mature (MII) eggs, number of frozen eggs, number of thawed eggs (e.g., eggs that survive thawing), number of embryos, number of blastocysts, number of euploid blastocysts, number of usable blastocysts / euploid blastocysts) treatment cycle number (i.e., retrieval cycle number or insemination cycle number based on prior and / or predicted number of retrieval cycles or IUI cycles, respectively), transfer number from retrieval cycle, PGT status of one or more embryos (e.g., blastocysts or euploid blastocysts), storage status of one or more eggs and / or embryos (e.g., fresh or frozen), race, ethnicity, surgical history (e.g., prior uterine surgery history), medications, and measurements such as measurements of AMH, E2, LH, P4, FSH, and / or AFC, sperm parameters (e.g., concentration and / or motility), and / or the like.

[0081] In some variations, the PFP tool may be trained on a large dataset of prior patient data. For example, the dataset of prior patient data may include data from a plurality of prior patient treatment (e.g., retrieval or IUI) cycles. For example, the plurality of patient cycles may include at least 10 cycles, at least 50 cycles, at least 100 cycles, at least 500 cycles, at least 1,000 cycles, at least 2,000 cycles, at least 3,000 cycles, at least 4,000 cycles, at least 5,000 cycles, at least 6,000 cycles, at leastATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 7,000 cycles, at least 8,000 cycles, at least 9,000 cycles, at least 10,000 cycles, at least 15,00 cycles, at least 20,000 cycles, at least 25,000 cycles, at least 30,000 cycles, at least 40,000 cycles, at least 50,000 cycles, at least 60,000 cycles, at least 70,000 cycles, at least 75,000 cycles, at least 100,000 cycles, at least 125,000 cycles, at least 150,000 cycles, at least 175,000 cycles, at least 200,000 cycles, at least 250,000 cycles, or at least 500,000 cycles. In some variations, the PFP tool may include a plurality of models, and each of the models may be trained on unique datasets. For example, a first set of the plurality of models may be trained on a dataset comprising thousands to hundreds of thousands of publicly accessed prior patient cycles (e.g., from a public database like SART CORS), while a second, different set of the plurality of models may be trained on a dataset comprising thousands to hundreds of thousands of privately collected prior patient cycles. For example, one or more of the models may be trained on prior patient data collected from one or more medical establishments. In some variations, one or more of a plurality of models of the PFP tool may not be trained with a dataset.

[0082] In some variations, one or more models of the PFP tool may be retrained periodically (e.g., at set intervals such as one a month, once every six months, once every year, etc.) or nonperiodically (e.g., at non-set intervals) using patient data acquired between the previous model training and the current model training. For example, one or more models of the PFP tool may be retrained using one or more prior egg outcomes and / or fertility outcomes for a patient. In some variations, one or more of the PFP tools described herein may be trained using prior patient data collected by one or more medical establishments (e.g., by a single medical establishment) during the year prior to model employment at the one or more medical establishments. Alternatively, in some variations, the one or more models may be continuously retrained (e.g., retrained whenever new patient data is acquired). Further, in some variations, the one or more models may be retrained upon detection of model decay (e.g., concept drift and / or data drift). For example, when signs of model decay are observed or detected, recently collected data (e.g., collected within about 1 month to about 6 months prior to the detected model decay) may be used to retrain the model, thus improving the accuracy of model outputs.

[0083] In some variations, the predictive family planning tool may be embodied as a non- transitory computer-readable medium storing instructions for execution by one or more processors. The processors may be communicably coupled to an electronic medical record (EMR) database forATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 receiving patient data. This matches both claims almost word-for-word.

[0084] The PFP tool may provide improvements to clinical outcomes by facilitating earlier and more accurate treatment scheduling. For example, when a patient receives a prediction that their probability of achieving multiple live births decreases significantly with each year of delay, the patient and medical professional may elect to initiate ovarian stimulation earlier, thereby preserving higher quality oocytes. Similarly, when a report indicates a higher-than-expected likelihood of blastocyst attrition, a physician may elect to increase the number of retrieval cycles in the initial treatment plan rather than after failed transfers, thereby avoiding wasted cycles and reducing patient stress. These improvements to treatment planning and timing thus improve conventional counseling tools, which typically provide static cumulative live birth estimates that are not individualized to multi-child family goals.

[0085] Additionally, use of the PFP tool may reduce the number of unnecessary treatment cycles, thereby decreasing patient exposure to invasive procedures and prolonged hormone therapy. This in turn may also reduce clinic resource utilization such as embryology lab and operating room use. These reductions in medical procedures and resource consumption also distinguish the predictive family planning tool from conventional abstract prediction tools, as they reflect measurable improvements to both the functioning of fertility clinics and patient health outcomes.

[0086] Variations of models of the PFP tool are described below and may include: an egg outcome (EO) model, an egg outcome attrition model, a first fertility outcome success model, a fertility attrition model, a fertility outcome failure model, and a second fertility outcome success model. In some variations, the PFP tool may comprise a variation of each of at least the egg outcome, egg outcome attrition, and first fertility outcome success models. 1.1. Egg Outcome Model

[0087] The egg outcome (EO) model may be configured to predict an egg outcome (or range thereof) for a retrieval cycle based on patient data. In some variations, the EO model may be trained on the order of thousands, tens of thousands, or hundreds of thousands of prior patient cycles retrieved from a database (e.g., a public database like SART CORS). In some variations, theATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 predicted egg outcome may be received by another model of the PFP tool (e.g., the egg outcome attrition model) to generate another prediction.

[0088] FIG.2A depicts an exemplary configuration of an EO model 200. The input 202 to the EO model 200 may include patient data, such as one or more of the aforementioned datapoints. For example, the patient data may include one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). The EO model 200 may be any of the models described herein, such as, for example, a regression model. In some variations, the EO model 200 may be a linear model, such as a Poisson regression model. The output 204 of the EO model 200 may be a predicted egg outcome or predicted range for the egg outcome. In some variations, the egg outcome may be a number of oocytes, a number of MII eggs, a number of fertilized eggs, a number of frozen eggs, a number of zygotes (i.e., 2PNs), etc.

[0089] The EO model may be run any number of times to provide a predicted egg outcome for each of a corresponding number of treatment cycles. The patient data input into the EO model may be adjusted to achieve this. For example, treatment cycle number may be adjusted to predict an egg outcome for a retrieval cycle that may occur after a first treatment cycle. Furthermore, one or more variations of the EO model may be employed by the PFP tool. For example, a first EO model may be configured to predict a number of oocytes (or range thereof) retrieved from the patient. A second EO model may be configured to predict a number of mature MII eggs (or range thereof) for the patient. A third EO model may be configured to predict a number of zygotes (2PNs) (or range thereof) for the patient. A fourth EO model may be configured to predict a number of frozen eggs (or range thereof) for the patient. A fifth EO model may be configured to predict a number of thawed eggs (or range thereof) for the patient. One or more of the first through fifth models may be implemented so that the resultant predicted egg outcomes may be provided to a patient, such as, for example, on a family planning report. For example, one or more of these predicted egg outcomes, or a range thereof, may be provided along a fertility treatment timeline and / or egg / embryoATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 development to inform a patient as to when each egg outcome will be realized during the treatment and how the value or range of values for the egg outcome changes as the treatment continues. It should be understood that all of the first through fifth exemplary EO models need not be run to provide family planning predictions and / or reports to patients. Similarly, it should be understood that, when more than one of the first through fifth exemplary EO models are run to provide family planning predictions, they need not be executed in any particular order. Further, in some variations, a single model may be configured to predict more than one egg outcome. For example, a single model may be configured to predict a number of frozen eggs and a number of thawed eggs for a patient.

[0090] With respect to the fourth and fifth (frozen and thawed eggs) EO model variations, the EO models may be implemented as one model or separate models to adjust a predicted egg outcome to account for expected losses throughout a treatment cycle (or at least a portion thereof). For example, following egg retrieval, a portion of the oocytes retrieved may not develop into MII eggs, and thus may be discarded instead of frozen. Following egg freezing, a portion of the MII eggs that were frozen may survive thawing. Accordingly, for patients planning to or considering egg freezing, the EO model may adjust the predicted egg outcome by applying a retention factor to the predicted egg outcome. The retention factor may account for expected losses at particular biological checkpoints. For example, the retention factor may include one or both of a maturity retention factor and a thaw retention factor, each of which may be predetermined (e.g., based on historical data, such as studies and / or prior patient data). FIG.2B depicts an exemplary implementation of the EO model 200, which is configured to apply a retention factor to the predicted egg outcome, resulting in an adjusted egg outcome.

[0091] The maturity retention factor may account for a percentage of the egg outcome that is predicted to be lost during maturation. For example, the maturity retention factor may represent a percentage of retrieved oocytes expected to reach mature stage suitable for freezing. The percentage may be about 5% to about 50%, such as about 10% to about 45%, about 15% to about 40%, about 20% to about 35%, or about 35% to about 30% (including all ranges and subranges therebetween). In some variations, the percentage of the egg outcome predicted to be lost during maturation may be about 10%, about 15%, about 20%, about 25%, about 30%, or about 35% (e.g., 20%, 25%, or 30%). To account for this percentage, the maturity retention factor may be a percentage of about 50% toATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 about 95%, such as about 55% to about 90%, about 60% to about 85%, about 65% to about 80%, or about 70% to about 75% (including all ranges and subranges therebetween). In some variations, the maturity retention factor may be a percentage of about 65%, about 70%, about 75%, about 80%, about 85%, or about 90% (e.g., 70%, 75%, or 80%).

[0092] Moreover, the thaw retention factor may account for a percentage of the egg outcome that is predicted to be lost during thawing. Put differently, the thaw retention factor may represent a percentage of frozen eggs expected to survive the thawing process when needed for fertilization. The percentage may be about 0% to about 25%, such as about 1% to about 20%, about 5% to about 15%, about 7% to about 12%, or about 9% to about 11% (including all ranges and subranges therebetween). In some variations, the percentage of the egg outcome predicted to be lost during thawing may be about 1%, about 3%, about 5%, about 7%, about 9%, about 10%, about 11%, about 13%, about 15%, about or 20% (e.g., 5%, 10%, or 15%). To account for this percentage, the thaw retention factor may be a percentage of about 75% to about 100%, such as about 80% to about 99%, about 85% to about 95%, about 88% to about 93%, or about 89% to about 91% (including all ranges and subranges therebetween). In some variations, the thaw retention factor may be about 80%, about 85%, about 87%, about 89%, about 90%, about 91%, about 93%, about 95%, about 97%, or about 99% (e.g., 85%, 90%, or 95%).

[0093] In some variations, one or both of the maturity and retention factors may be selected or adjusted by a medical professional, or by the EO model. In some variations, the second (MII eggs) EO model may be implemented to predict a number of mature eggs for a patient, which may be substantially equivalent to predicting a number of eggs (e.g., number of oocytes) with the maturity and retention factors via the first EO model. For example, an EO model that predicts a number of mature (MII) eggs may additionally be configured to determine a retention factor for a patient based on prior patient data, and may be configured to combine a predicted number of eggs for the patient with the determined retention factor to provide a predicted number of MII eggs for the patient.

[0094] For patients who do not have eggs frozen from previous retrieval cycles, the EO model may be configured to apply one or both of the maturity and thaw retention factors to the predicted egg outcome (or range thereof) for each of a plurality of retrieval cycles taken into account by the PFP tool. Alternatively, for patients who do have eggs from one or more previous retrieval cycles,ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 the EO model may be configured to predict and adjust, via one or both of the maturity and thaw retention factors, the egg outcome for one or more future retrieval cycles for the patient.

[0095] In some variations, one or more egg outcomes from the EO model may be provided to a patient (e.g., via a family planning report or other suitable format) so that the patient may have a comprehensive understanding of an overall fertility treatment cycle. For example, for an IVF cycle, one or more variations of the EO model may be implemented to provide one or more egg outcomes each associated with a particular step of a retrieval cycle (e.g., egg retrieval, egg freezing, egg thawing, egg fertilization, embryo maturation, etc.) so that the patient has context for a predicted success rate of their desired fertility outcome. That is, the patient may better understand the predicted success rate of their desired fertility outcome if they are aware of how their egg outcome is predicted to change (e.g., decline) throughout the cycle. 1.2. Egg Outcome Attrition Model

[0096] The egg outcome attrition (EOA) model may be configured to predict a probability distribution of an egg outcome based on patient data and a predicted egg outcome. In some variations, the EOA model may be configured to receive the predicted egg outcome from the EO model, which may be a first egg outcome, to predict a probability distribution for a second, different egg outcome that develops from the first egg outcome. For example, the first egg outcome may be a predicted a number of eggs, and the EOA model may predict a probability distribution for a number of embryos or blastocysts (e.g., usable blastocysts, euploid blastocysts) that will develop from the predicted number of eggs. To do so, the EOA model may predict, using patient data, an attrition rate (%) that accounts for loss of the first egg outcome through egg and / or embryo development. Next, the EOA model may combine the predicted attrition rate and the predicted first egg outcome to yield a probability distribution for the second egg outcome. In some variations, the predicted probability distribution may be received by another model of the PFP tool (e.g., the first fertility outcome success model) to generate another prediction.

[0097] In some variations, the EOA model may be trained on the order of thousands, tens of thousands, or hundreds of thousands of prior patient cycles retrieved from a database (e.g., a public database like SART CORS).ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0098] FIG.3 depicts an exemplary implementation of an EOA model 300. A first input 302 to the EOA model 300 may include patient data, such as one or more of the aforementioned datapoints. For example, the patient data may include one or more of: age (e.g., at time of given retrieval cycle) BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). In some variations, the patient data may include only one datapoint (e.g., patient age at time of given retrieval cycle). A second input 303 to the EOA model 300 may be the output from an EO model (e.g., EO model 200), a predicted egg outcome (first egg outcome) or predicted range thereof. The EOA model 300 may include any of the models described herein, such as, for example, a regression model. In some variations, the EOA model 300 may include a logistic regression model. A first (intermediate) output 304 of the EOA model 300 may include an egg outcome attrition rate. The first output 304 may be combined with the second input 303 for the model 300 to generate a second (final) output 305, which may be a probability distribution of an egg outcome (second, different egg outcome that develops after the first egg outcome). In some variations, the second egg outcome may be a number of embryos (e.g., a number of blastocysts, such as usable or euploid blastocysts).

[0099] The EOA model may be executed any number of times to provide a predicted probability distribution for an egg outcome for each of a corresponding number of treatment cycles. The patient data input into the EOA model may be adjusted to achieve this. For example, patient age and may be adjusted to predict an egg outcome for a retrieval cycle that may occur in the future (e.g., in at least 6 months, at least 1 year, at least 2 years, etc.). In some variations, a plurality of predicted egg outcome probability distributions generated for each of a plurality of retrieval cycles may be combined into a single predicted probability distribution output.

[0100] Furthermore, one or more variations of the EOA model may be employed by the PFP tool. For example, a first EOA model may be configured to predict probability distribution for a number of usable blastocysts that will develop from a predicted number of eggs (e.g., oocytes) received from the EO model. As another example, a second EOA model may be configured to predictATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 probability distribution for a number of euploid blastocysts that will develop from a predicted number of eggs (e.g., oocytes) received from the EO model.

[0101] In some variations, a predicted probability distribution for an egg outcome may be provided to a patient (e.g., via the family planning report) so that the patient may make a more informed decision about when to begin fertility treatment. As a result, the patient may choose to adjust or schedule a particular start date for a treatment cycle. In some variations, a predicted probability distribution for an egg outcome may be provided (e.g., via an indicator, which may be a graph) to a patient on a family planning report. For example, for patients electing to use PGT during fertility treatment, a predicted probability distribution for euploid blastocysts may be available for review within the report.

[0102] FIG.4 illustrates the results of two exemplary EOA models. Plot 402 of FIG.4 shows how a euploid blastocyst attrition rate varies with patient age according to an implementation of a euploid attrition rate model. Plot 404 of FIG.4 shows how a usable blastocyst attrition rate varies with patient age according to an implementation of a blastocyst attrition rate model. Plots 402 and 404 reveal that older patients, such as patients over 35, may experience increased rates of attrition for for euploids and usable blastocysts. 1.3. First Fertility Outcome Success Model

[0103] The first fertility outcome success (FOS-1) model may be configured to predict a probability that one or a plurality of fertility outcomes will result from a treatment cycle for a patient based on patient data and a predicted probability of an egg outcome. In some variations, the fertility outcome(s) may include at least one fertility outcome, at least two fertility outcomes, at least three fertility outcomes, at least four fertility outcomes, at least five fertility, or five or more fertility outcomes. As noted above, a fertility outcome may include, for example, a live birth, a pregnancy, or an embryo transfer. For example, the fertility outcome(s) may include at least one live birth, at least two live births, at least three live births, at least four live births, or at least five live births. In some variations, the fertility outcome(s) may include any desired number of live births, which may be input into the PFP tool via the GUI. In some variations, the FOS-1 model may be configured to predict the chances of success for a range of numbers of fertility outcomes, such as a probability that each of one, two, and three fertility outcomes will result from a single retrievalATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 cycle. Providing this plurality of predicted success rates (e.g., via the family planning report) may anticipate and address patient questions related to their chances of having their desired number of children.

[0104] To predict a success rate for a fertility outcome, the FOS-1 model may be configured to predict, based on patient data, a first success rate for a first fertility outcome. The first fertility outcome may be related to the egg outcome for which a probability distribution was predicted by the EOA model. For example, the egg outcome may be a number of embryos (e.g., a number of blastocysts, such as euploid blastocysts or usable blastocysts), and the first fertility outcome may be a embryo transfer. The FOS-1 model may be configured to receive the predicted egg outcome probability distribution from the EOA model and combine the predicted success rate for the first fertility outcome with the predicted egg outcome probability distribution to determine a probability that one and / or a plurality of second, different fertility outcomes will result from the retrieval cycle. The second fertility outcome(s) may be, for example, live birth or pregnancy.

[0105] In some variations, the FOS-1 model may be trained on the order of thousands, tens of thousands, or hundreds of thousands of prior patient cycles retrieved from a database (e.g., a public database like SART CORS).

[0106] FIG.5 depicts an exemplary implementation of an FOS-1 model 500. A first input 502 to the FOS-1 model 500 may include patient data, such as one or more of the aforementioned datapoints. For example, the patient data may include, for example, one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). The FOS-1 model may include any of the models described herein, such as, for example, a regression model. In some variations, the FOS-1 model 500 may include a logistic regression model. The output 506 of the FOS-1 model 200 may be a fertility outcome success rate, which may be a first (intermediate) output. This output 506 may be applied to a second input 504 for the model 500, which may be a probability distribution for an egg outcome. In some variations, the model 500 may receive the inputATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 504 from another one of the PFP models, such as from the EOA model, or from a combination of the EO model and the EOA model. As a result, the FOS-1 model 500 may generate a second (final) output, which may be a probability that one and / or a plurality of second, different fertility outcomes will result from the retrieval cycle. The second fertility outcome(s) may be fertility outcome(s) that occur later in time than the first fertility outcome. For example, the first fertility outcome may be an embryo transfer or a pregnancy, and the second fertility outcome may be a pregnancy or a live birth.

[0107] Generally, the final predicted probability of success for the fertility outcome(s) may be provided to a patient, such as, for example, via the family planning report. For example, the predicted success rates for each of one fertility outcome, two fertility outcomes, and, optionally, three fertility outcomes (e.g., differentiated by a quantity of the fertility outcome) may be indicated on the family planning report. In some variations, the predicted success rate for at least two fertility outcomes (e.g., at least two live births), and optionally one and / or three fertility outcomes (e.g., one and / or three live births) may be presented as the patient^s percent chance of having (at least) 2 children as a result of the given treatment cycle, and optionally (at least) 1 child, and / or (at least) 3 children as a result of the given treatment cycle.

[0108] FIG.6 illustrates the implementation results and analysis of an exemplary FOS-1 model. Plot 602 of FIG.6 shows a probability density of live birth for a sample of patients given their predicted probability distributions for euploid blastocysts or untested, usable blastocysts. The plot 602 shows that the predicted probability of live birth is proportional to the density of the predicted blastocyst probability distribution. Moreover, plot 604 of FIG.6 depicts the correlation between the predicted probability of live birth and the actual live birth results from the sample. As shown, the correlation is strong, indicating that the exemplary FOS-1 model is highly calibrated. 1.4. Fertility Attrition Model

[0109] The fertility attrition (FA) model may be configured to predict how one or more patient fertility characteristics may decline over time based on patient data. The one or more patient fertility characteristics may include a subset of datapoints from the patient data described herein, and may include, for example, measurements of AMH, E2, LH, P4, FSH, and / or AFC. In some variations, the measurements may include a most recent measurement of AMH, E2, LH, P4, FSH, and / or AFC. In some variations, the measurements may include an average of a plurality of recent measurement ofATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 AMH, E2, LH, P4, FSH, and / or AFC. In some variations, the FA model may be configured to estimate a most-recent measurement (e.g., from a current day) of AMH, E2, LH, P4, FSH, and / or AFC for a patient based on a recent measurement (e.g., from one, two, three, four, five, or more than five days prior to the current day) of AMH, E2, LH, P4, FSH, and / or AFC.

[0110] FIG.7 depicts an exemplary implementation of an FA model 700. The input 702 to the FA model 700 may include patient data, such as one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). In some variations, the patient data may be variable (e.g., age after a predetermined time period, such as at least 6 months, at least 1 year, at least 2 years, at least 3 years, at least 4 years, or at least 5 years from the present). The FA model 700 may be any of the models described herein, such as, for example, a regression model. In some variations, output 704 of the FA model 700 may be a predicted level of AMH, E2, LH, P4, FSH, and / or AFC for the patient. In some variations, the output 704 of the FA model 700 may be a predicted rate of attrition for a level of one or more of AMH, E2, LH, P4, FSH, and / or AFC for the patient. This output 704 may be input as patient data into the EO, EOA, and FOS-1 models to provide family planning predictions with the predicted level(s) instead of the most recent level(s) thereof, or for determining a future level using the predicted attrition rate and one or more of the EO, EOA, and FOS-1 models.

[0111] In some variations, the FA model may be an AMH attrition model that predicts, using a patient age as an adjustable input, how the patient^s AMH level will decline over time. That is, the AMH attrition model may provide a predicted level of AMH for a given input age. As an example, FIG.8 depicts a plot 800 that shows how predicted AMH levels for sample patients decline as patient age increases.

[0112] In some variations, the FA model may be trained on the order of thousands, tens of thousands, or hundreds of thousands of prior patient cycles retrieved from a database (e.g., a public database like SART CORS).ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0113] In some variations, the FA model may be implemented along with the EO, EOA, and FOS- 1 models to provide for comparison, via, for example, the family planning report, the success rates for one or each of a plurality of fertility outcomes of at least one first retrieval cycle and at least one second retrieval cycle, where the first retrieval cycle occurs a predetermined amount of time (e.g., at least six months, at least one year, at least 1.5 years, at least 2 years, etc.) prior to the second retrieval cycle. This comparison may help a patient understand how age, by varying indirectly with AMH level, may result in different (e.g., lower) probabilities of success of the fertility outcome(s). Accordingly, the FA model may facilitate a patient^s decision regarding a timing of a fertility treatment (e.g., a next treatment cycle). For example, a patient may choose to schedule a start date for a fertility treatment, or adjust a planned start date, based on the FA model predictions. 1.5. Fertility Outcome Failure Model

[0114] The fertility outcome failure (FOF) model may be configured to predict a probability of failure of a fertility outcome for a patient based on patient data. The failed fertility outcome may include, for example, pregnancy loss in the form of miscarriages, abortions, and / or still births following embryo transfer or successful IUI. In some variations, the FOF model may be implemented a first time to predict a first failure rate for a fertility outcome and a second time to predict a second failure rate for a fertility outcome. The first failure rate may be determined based on a planned or actual PGT status (e.g., untested or tested) of one or more of the patient^s embryos, and the second failure rate may be determined based on the opposite PGT status (e.g., tested or untested) for the one or more embryos. Accordingly, the FOF model may provide for comparison a first predicted failure rate for a fertility outcome based on the planned or actual PGT status for the patient and a second predicted failure rate based on the opposite PGT status for the patient. These predictions, when provided together, may illustrate for a patient how PGT testing impacts their chances of failure (e.g., miscarriage) of a fertility outcome (e.g., embryo transfer). Additionally, or alternatively, in some variations, the first predicted failure rate may be further based on an actual or planned storage status (e.g., fresh or frozen) for the one or more embryos of the patient, and the FOF model may be implemented a third time to predict a third failure rate for a fertility outcome. The third predicted failure rate may be determined based on the opposite embryo storage status (e.g., frozen or fresh), and may be provided to the first failure rate so that the patient mayATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 understand how embryo storage impacts their chances of failure (e.g., miscarriage) of a fertility outcome (e.g., embryo transfer).

[0115] In some variations, the FA model may be trained on the order of thousands, tens of thousands, or hundreds of thousands of prior patient cycles retrieved from a database (e.g., a public database like SART CORS).

[0116] FIG.9 depicts an exemplary implementation of an FOF model 900. The input 902 to the FOF model 900 may include patient data, such as one or more of the aforementioned datapoints. For example, the patient data may include one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). The model 900 may receive, for example, at least a PGT status of one or more embryos and / or a storage status of one or more embryos. The FOF model 900 may be any of the models described herein, such as, for example, a regression model. The output 904 of the FOF model 900 may be a predicted probability for one a fertility outcome(s). This output 904 may be provided to the patient (via, e.g., a family planning report) to illustrate how PGT may lower risk of failed fertility outcomes, and especially for older patients (e.g., patients older than 35).

[0117] FIG.10 illustrates the implementation results and accuracy of an exemplary FOF model. Plot 1002 of FIG.10 shows how predicted probability of pregnancy loss (in this example, failure due to miscarriage and abortion) generally increases with patient age. In particular, this predicted risk is significantly mitigated for embryo transfers using frozen, tested (i.e., euploid) embryos as compared to the risk for embryo transfers using fresh and frozen, untested embryos. Moreover, plot 1004 of FIG.10 depicts the correlation between the predicted probability of pregnancy loss and the actual pregnancy loss results for the sample. As shown, the correlation is strong, indicating that the exemplary FOF model is highly calibrated.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 1.6. Second Fertility Outcome Success Model

[0118] The second fertility outcome success (FOS-2) model may be configured to predict a probability of successful fertility outcome for a patient based on patient data, where the fertility outcome may result from an IUI cycle. The successful fertility outcome may include, for example, live birth, pregnancy, or embryo transfer. In some variations, the FOS-2 model may be trained on the order of hundreds, thousands, tens of thousands, or hundreds of thousands of prior patient cycles of a dataset (e.g., a dataset built and stored on a private database).

[0119] FIG.12 depicts an exemplary implementation of an FOS-2 model 1200. The input 1202 to the FOS-2 model 1200 may include patient data, such as one or more of the aforementioned datapoints. For example, the patient data may include one or more of: sperm parameters, age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). In some variations, the patient data may include at least a sperm parameter. The sperm parameters may include sperm concentration and / or sperm motility. In some variations, the FOS-2 model 1200 may be a regression model. The output 1204 of the FOS-2 model 1200 may be a predicted probability that one and / or a plurality of fertility outcomes may result from a treatment cycle, such as an insemination cycle. This output 1204 may be provided (e.g., via a family planning report) for a patient so that the patient may compare predicted success rates of IVF fertility outcomes and IUI fertility outcomes, allowing the patient to make an informed decision about which may be their preferred treatment. Accordingly, the FOS-2 model may facilitate a patient^s decision regarding a type of fertility treatment to pursue. 2. Family Planning Report

[0120] As discussed above, the technology described herein may be provided to patients and / or medical professionals to guide fertility consultations and inform decisions for fertility treatment. The models disclosed herein may provide interpretable predictions, such as in the form of a familyATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 planning report, via, for example, a GUI (e.g., user interface 110 of FIG.1, which may be provided by display 108). The GUI may be interactive in that it may provide one or more inputs configured to receive patient data for inputting into the PFP tool, receive a desired fertility outcome for inputting into the PFP tool, selecting a tool, adjust a setting, recalculate family planning predictions provided on the report, and / or the like.

[0121] In general, the family planning reports herein that are configured for viewing may include: (1) one or more report portions and / or (2) one or more user input portions for receiving patient data and patient family goals, to toggle between settings, and the like. The report portion(s) may include one or more prediction indicators, each for a family planning prediction, and / or one or more explanation indicators. The prediction indicators may include one or more of text (e.g., numbers and / or letters), an image, a symbol, and a graph that represents a prediction from the PFP tool. The explanation indicator(s) may provide context for the prediction indicators (but may not directly represent a prediction) in the form of text, images, or symbols (e.g., a timeline). In some variations, a first prediction indicator may represent a probability that a patient^s desired fertility outcome will result from each of one or more treatment cycles, a second indicator may represent a fertility treatment (e.g., IVF or IUI) in the form of a plurality of steps of the given treatment, and a third indicator may represent a predicted egg outcome for one of the plurality of steps of the treatment. The report portions and user input portions provided on a family planning report may be organized in any suitable position relative to one another. For example, a given report portion may be positioned above, below, or next to a user input portion, and / or above, below, or next to another report portion.

[0122] The GUI providing the family planning reports may facilitate real-time exploration of treatment scenarios through interactive elements. For example, adjustments to input parameters may trigger recalculation of predictions, allowing users to explore how factors such as treatment timing, number of cycles, or use of PGT affect outcome probabilities. As another example, predicted probabilities may be provided as graphics and / or with explanations, converting complex statistical outputs into actionable insights.

[0123] FIG.12A-FIG.19 described below illustrate exemplary variations of a GUI dashboard for providing a family planning report. While particular features of the GUI dashboards below may beATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 shown and described, it should be understood that a family planning report may be provided via a GUI that comprises fewer, or more, than all of these features.

[0124] FIG.12A depicts an exemplary GUI dashboard 1200, on which a family planning report for an IVF patient is provided. The dashboard 1200 may be configured to include a first user input portion 1202 that can be used, for example, to receive patient data, patient family goals, and / or to recalculate family planning predictions provided in the report. The dashboard 1200 may also be configured to include a second user input portion 1204 that may be used to receive additional patient data (e.g., PGT status) and to allow the family planning report to be shared, printed, or annotated by a user. The family planning report provided by the dashboard 1200 may include a first report portion 1206 and a second report portion 1210. The first report portion 1206 may have prediction indicators, including, for each of three retrieval cycles, an indicator 1207^ that shows the patient^s predicted chance of having a desired number of children (three), and / or indicators 1207^ and 1207^^ that show the patient^s predicted chance of having a number of children different from the desired number (one and two). The first report portion 1206 may also include explanation indicators 1209^ and / or 1209^ that provide context for the family planning predictions. Further, additional explanation indicators 1209^^ and 1209^^ may be configured to be selected in order to provide additional information. For example, upon selecting the indicator 1209^^, a third report portion (not shown) may be revealed to explain the family planning prediction calculations. Moreover, selecting the indicator 1209^^ may reveal an additional prediction / explanation indicator 1208. For example, referring briefly to FIG.12B, which shows a portion of the GUI dashboard 1200, the prediction / explanation indicator 1208 may represent a predicted probability distribution for an egg outcome of one or more of the retrieval cycles, along with an explanation of the predicted probability distribution.

[0125] Referring again to FIG.12A, the second report portion 1210 may include two explanation indicators 1211^ and 1211^. The explanation indicator 1211^ may show a timeline of the steps of an IVF retrieval cycle, with each step configured to be selected to reveal a corresponding optional explanation indicator (not shown) that may provide information about each step. The second report portion 1210 may also include one or more prediction indicators 1212^, 1212^, 1212^^, and 1212^^ that represent a predicted egg outcome from the PFP tool. In particular, prediction indicators 1212^ and 1212^ may represent ranges for predicted egg outcomes from the EO model, and predictionATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 indicators 1212^^ and 1212^^ may represent ranges for predicted egg outcomes from the EO model. The prediction indicators 1212^^ and 1212^^ may also be actionable indicators that, when selected, provide additional explanation indicators (not shown) to provide context for each prediction. Together, the prediction indicators 1212^, 1212^, 1212^^, and 1212^^ may make up a funnel prediction indicator that shows, in conjunction with the timeline of fertility steps, how profound egg outcome attrition may be throughout the retrieval cycle.

[0126] FIG.13A depicts an exemplary GUI dashboard 1300, on which a family planning report for an IVF patient considering egg freezing is provided. The dashboard 1300 may include many of the same or similar features as the dashboard 1200, such as the first and second user input portions 1302 and 1304, and the first and second report portions 1306 and 1310. Here, the first report portion 1306 may include prediction indicators including, for each of three retrieval cycles, an indicator 1307^ that shows the patient^s predicted number of eggs (e.g., oocytes) frozen, and / or an indicator 1307^ that represents the patient^s predicted chance of having one, two, and three children. Further, the first report portion 1306 may include a tool 1308 that is configured to be selected to enable a comparison of family planning predictions for the patient at a current age and a future age. Turning briefly to FIG.13B, which shows the first report portion 1306 of the GUI dashboard 1300, the comparison is provided via prediction indicators, including, for each of three retrieval cycles, indicators 1309^ that show the patient^s predicted number of eggs (e.g., oocytes) frozen, as well as, for a first retrieval cycle, indicators 1309^ that represent the patient^s predicted chance of having one and two children. As shown, the patient^s predicted egg outcome and predicted chances of achieving the desired number of children decline as patient age increases. This decline may be attributed to hormone (e.g., AMH) attrition rates and / or egg quality attrition rates, as provided by the FA model. Referring again to FIG.13A, because the dashboard 1300 is configured for a patient considering egg freezing, the explanation and prediction indicators 1311 and 1312 may be expanded (from those of FIG.12A) to represent predicted egg outcomes and treatment cycle steps for freezing and thawing eggs.

[0127] FIG.14 depicts an exemplary GUI dashboard 1400, on which a family planning report for an IVF patient who has already completed at least one retrieval cycle is provided. Here, the first user input portion 1402 may include a tool 1403 that is configured to be selected to import the patient^s data from the prior retrieval. Further, the first report portion 1406 may include, for aATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 current cycle, an indicator 1407 that shows the total number of frozen eggs from prior retrieval(s), and / or, for an additional cycle, a prediction indicator 1408 that represents the patient^s predicted total number of frozen eggs (i.e., the number of prior frozen eggs and the predicted number of frozen eggs). For both cycles, the first report portion 1406 may include prediction indicators 1409 that represent the patient^s predicted chance of having one, two, and three children.

[0128] In some variations, the family planning reports herein may include a report portion for providing information about PGT. Such an exemplary report portion 1500 is depicted in FIG.15. As shown, the report portion 1500 may include an explanation indicator 1502 and a prediction indicator 1504. The prediction indicator 1504 may include a first indicator 1505^ that represents the predicted probability of embryo transfer success and / or a second indicator 1505^ that represents the predicted probability of pregnancy loss following embryo transfer for PGT (euploid) embryos versus untested embryos. These family planning predictions may be provided at least in part by the FOF model described herein. The report portion 1500 may be provided on an individual dashboard of the GUI, provided adjacent to one of the first and second report portions described above, or selectively provided via an actionable tool on the main dashboard (i.e., the dashboard providing the first and second report portions described above).

[0129] In some variations, the family planning reports herein may include a report portion for providing information about IUI. Such an exemplary report portion 1600 is depicted in FIG.16. As shown, the report portion 1600 may include an IVF report portion 1602 and an IUI report portion 1604. The IVF report portion 1602 may include prediction indicators 1603 that each represent a predicted probability that a desired fertility outcome (e.g., desired number of live births) will result from a given retrieval cycle. The IUI report portion 1604 may include prediction indicators 1605, which represent a predicted probability of pregnancy for each of three IUI cycles, prediction indicators 1606, which represent a predicted probability of live birth for each of three IUI cycles, and / or an explanation indicator 1608 that provides information to explain the predictions indicated by the indicators 1605 and 1606. The report portion 1600 may be provided on an individual dashboard of the GUI, provided adjacent to one of the first and second report portions described above, or selectively provided via an actionable tool on the main dashboard (i.e., the dashboard providing the first and second report portions described above).ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030

[0130] In some variations, the family planning reports herein may include a report portion for providing family planning predictions based on patient data from a patient^s egg donor. Such an exemplary report portion 1700 is depicted in FIG.17. As shown, the report portion 1700 may include a donor report portion 1702 and a patient report portion 1704. Both the donor and patient report portions 1702, 1704 may include, for each of three retrieval cycles, a prediction indicator 1706 that represents a predicted chance of having a desired number of children. Both the donor and patient report portions 1702, 1704 may also include a prediction indicator 1708 that represents the predicted number of euploid embryos per cycle. In some variations, the report portion 1700 may allow a patient to make a more informed and / or quicker decision about whether to use their own eggs or donor eggs during a fertility treatment. The report portion 1700 may be provided on an individual dashboard of the GUI, provided adjacent to one of the first and second report portions described above, or selectively provided via an actionable tool on the main dashboard (i.e., the dashboard providing the first and second report portions described above).

[0131] In some variations, the family planning reports herein may include one or more report portions for providing treatment predictions based on the family planning predictions. The treatment predictions may be based at least in part on predicted egg outcomes from the EO model. The treatment predictions may include, for example, a predicted medication dose (e.g., of ovarian stimulation medication) for the patient, and / or a predicted hormonal trigger day for the patient. For example, as shown in FIG.18, an optimal starting dose report portion 1800 may include a prediction indicator 1802 representing a plurality of predicted starting doses and identifying an optimal dose for the patient within the plurality, as well as an explanation indicator 1804 for explaining the predictions provided via the indicator 1802. As another example, as shown in FIG.19, an optimal trigger day report portion 1900 may include a prediction indicator 1902 representing a plurality of predicted trigger days and identifying an optimal trigger day for the patient within the plurality, as well as an explanation indicator 1904 for explaining the predictions provided via the indicator 1902. In some variations, a patient may be administered medication (e.g., a starting dose of ovarian stimulation medication, and / or a hormonal trigger shot) based on the family planning predictions noted in the report portions 1800 and / or 1900. Additional systems, models, and methods for providing optimal dose and trigger day predictions, and aspects thereof, may be provided in U.S.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 Pat. No.11,735,302, and U.S. Pat. App. Serial No.18 / 613,016, the content of each of which was previously incorporated by reference herein.

[0132] While variations of family planning reports are described above, it should be understood that the family planning reports may be provided in any suitable format and design. In some variations, a family planning report may provide family planning predictions for each of one, two, three, four, five, six, seven, eight, nine, ten, or more than ten treatment cycles for a patients. Additionally, or alternatively, in some variations, the report portions of the family planning report may be shown in a different order, such as side by side or with the second report portion above the first report portion (e.g., higher on a page or dashboard providing the report). II. Methods for Family Planning

[0133] Methods for family planning using the predictive family planning tool are described in detail below. The methods herein may yield family planning predictions (e.g., success rate for fertility outcome and / or predicted egg outcome) and / or reports for review by a patient. The methods may optimize a fertility treatment for patients by facilitating fertility counseling, educating patients about numerous treatment options, and predicting success rates for patient-specified family goals, thus improving the field of ART. For example, any one of the methods herein may include a step for determining and / or adjusting a treatment plan for a patient based on one or more family planning predictions (and optionally a family planning report). For example, a patient and / or medical professional may use one or more family planning predictions to inform decisions related to one or more of: a type of ART treatment (IVF, IUI), a treatment protocol (e.g., number of IVF retrieval cycles, a number of egg freezing cycles, a number of IUI cycles, a number of embryo biopsies, a number of embryo transfers), use of preimplantation genetic testing (PGT), use of donor eggs, use of donor sperm, a medication dosage (e.g., of ovarian stimulation medication), a hormonal trigger day, and / or the like. The family planning prediction(s) and / or report may be used to adjust a planned fertility treatment for a patient, such as to adjust one or more of: a treatment timing (e.g., accelerate or delay a start date for a given treatment), a treatment protocol (e.g., increase or decrease a number of treatment cycles), a medication dosage (e.g., increase or decrease a starting dosage of ovarian stimulation medication), a hormonal trigger day timing (e.g., accelerate or delay the hormonal trigger day). As another example, based on the family planning prediction(s) and / or report, theATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 patient may elect to use one or more of preimplantation genetic testing (PGT), donor eggs, and donor sperm to increase their chances of achieving their desired fertility outcomes.

[0134] Put another way, in some variations, determining the fertility treatment plan may include one or more of: determining a type of fertility treatment for the patient, determining or adjusting a number of treatment cycles for the patient, determining or adjusting a start date for a treatment cycle for the patient, determining an egg freezing protocol for the patient, determining an embryo freezing protocol for the patient, determining a PGT protocol for the patient, and determining an egg type for the patient. In some variations, determining the type of fertility treatment may include electing an in vitro fertilization (IVF) process or an intrauterine insemination (IUI) process. In some variations, determining or adjusting the number of treatment cycles may include increasing or decreasing a number of treatment cycles. In some variations, determining or adjusting the start date for the treatment cycle may include advancing or delaying the start date. In some variations, determining the egg freezing protocol may include electing to freeze or to not freeze one or more eggs of the patient. In some variations, determining the embryo freezing protocol may include electing to freeze or to not freeze one or more embryos of the patient. In some variations, determining the PGT protocol may include electing to test or to not test one or more embryos of the patient. In some variations, determining the egg type may include electing to use eggs of the patient or eggs of a donor during fertility treatment.

[0135] Similarly, in some variations, any of the methods herein may include a step for treating a patient based on one or more family planning predictions. Treating the patient may include administering medication, such as ovarian stimulation medication or a trigger shot, and / or performing a procedure, such as egg retrieval, embryo biopsy, embryo transfer, and / or the like in order to support an optimized fertility process for the patient.

[0136] The methods may be computer-implemented (e.g., via a system such as system 100 of FIG.1) to execute the PFP tool (e.g., one or more models thereof). In some variations, any of the models herein may include a step for receiving patient data for a patient of interest. The patient data may be received from a database (e.g., an EMR) and / or from user input, such as via a display (e.g., touchscreen) or other input tool (e.g., mouse, keyboard, button, etc.) coupled thereto). The display may support a GUI configured to request and receive the user input. Additionally, in some variations, any of the models herein may include a step for receiving a desired fertility outcome forATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 a patient (e.g., via user input). In some variations, any of the models herein may include a step for training the model with prior patient data (and / or retraining the model with prior patient data and / or patient data for the patient of interest).

[0137] While particular steps of exemplary methods may be described in a particular order, it should be understood that, in some variations, one or more of the steps may rearranged within the method order, may be repeated any suitable number of times, or may be optional, and the methods may include feedback loops and / or additional steps.

[0138] FIG.20 depicts a flow diagram of a method 2000 for optimizing fertility treatment for a patient using a predictive family planning tool. In some variations, the fertility treatment may be an IVF treatment. At step 2002, the method 2000 may include predicting a probability of success for a fertility outcome of a retrieval cycle for the patient based on one or more models having received patient data. The patient data may comprise, for example, one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos (e.g., tested or untested). In some variations, the patient data may include infertility diagnosis, transfer number from treatment cycle, and / or one or more additional patient datapoints (e.g., at least infertility diagnosis and number of prior retrieval cycles). The one or more models may be configured to receive the patient data from an EMR and / or via patient input to a GUI configured to receive the input. The one or more models may comprise one or more models of the PFP tool, such as at least the FOS-1 model. Next, at step 2004, the method 2000 may include providing (e.g., via a family planning report) an indicator of the predicted probability of success for the fertility outcome. In some variations, the fertility outcome may be a plurality of fertility outcomes, such as at least two or at least three live births or pregnancies resulting from the retrieval cycle. In some variations, the method 2000 may be rerun any suitable number of times to make predictions for a corresponding number of treatment cycles.

[0139] FIG.21 depicts a flow diagram of another method 2100 for optimizing fertility treatment for a patient using a series of models of a predictive family planning tool. In some variations, theATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 fertility treatment may be an IVF treatment. At step 2102, the method 2100 may comprise predicting a probability distribution for an egg outcome of a retrieval cycle for the patient based on a first model having received first patient data. In some variations, the egg outcome may be a number of oocytes, a number of zygotes, a number of frozen eggs, a number of thawed eggs, or a number of mature (MII) eggs. In some variations, the method 2100 may include predicting the egg outcome for use by the first model, such as predicting an egg outcome via an EO model. In some variations, the step 2102 may be executed by an EOA model. Accordingly, in some variations, the step 2102 may additionally include predicting an attrition rate based on the first model (e.g., an EOA model) and the first data. Next, the method 2100 may include step 2104 for predicting a success rate for a first fertility outcome for the patient based on a second model having received second patient data. In some variations, the first and second data may include one or more of the same or different datapoints. For example, the first and second patient data may each comprise one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos.. Similarly, in some variations, the first and second models may be trained on first and second sets of prior patient data. The first and second sets of prior patient data may comprise overlapping datapoints, nonoverlapping datapoints, or all of the same datapoints. In some variations, the first fertility outcome may be an embryo transfer. Accordingly, in some variations, the probability of success for the first fertility outcome may be a pre-transfer embryo success rate, based on the second model and the second patient data. In some variations, the second model may be a FOS-1 model.

[0140] Next, at step 2106, the method 2100 may include generating a probability of success for a second fertility outcome based on the egg outcome probability distribution and the success rate for the first fertility outcome. In some variations, the first and second fertility outcomes may be different. For example, as noted above, the first fertility outcome may be an embryo transfer while the second fertility outcome may be live birth, such as at least one, at least two, or at least three live births resulting from the treatment cycle. Finally, the method 2100 may include step 2108 forATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 providing an indicator of the probability of success for the second fertility outcome via, for example, a report. The report may be a family planning report that is provided in the form of, for example, a verbal report (e.g., via a medical professional), a printed report, and / or an electronic report (e.g., a PDF, an email and / or via a graphical user interface (GUI) provided via a display). In some variations, the method 2100 may be rerun any suitable number of times to make predicts for a corresponding number of treatment cycles.

[0141] In some variations, the method 2100 may include predicting a fertility attrition rate for a patient based on patient data, such as via the FA model described herein. Further, the predicted fertility attrition rate may be used to predict a probability of a desired fertility outcome for the patient following a predetermined time period, such as after at least 6 months, after at least 1 year, after at least 1.5 years, after at least 2 years, etc. The desired fertility outcome may be the second fertility outcome provided by the step 2108. Next, the method 2100 may additionally include providing an indicator of the probability of the desired fertility outcome for the patient following the predetermined time period (e.g., via the family planning report).

[0142] Additionally, or alternatively, in some variations, the method 2100 may include predicting a probability of failure for a desired fertility outcome, such as via the FOF model described herein. In some variations, the method 2100 may additionally include of the failure rate for the desired fertility outcome for the patient (e.g., via the family planning report).

[0143] FIG.22 depicts a flow diagram of a method 2200 for graphically representing family planning predictions from a predictive family planning tool. To begin, the method 2200 may include receiving a desired fertility outcome and patient data, which may be achieved at a processor. The patient data may include, for example, one or more of: age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos.

[0144] At step 2204, the method 2200 may include determining a probability that the desired fertility outcome will result from each of one or more retrieval cycles for the patient based on theATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 patient data. Step 2204 may be achieved via one or more models of the PFP, such as via a sequence including at least the EO, EOA, and FSO-1 models. In some variations, each model may utilize a unique subset of the received patient data to contribute to predicting the probability of success for the desired fertility outcome. Step 2206 may include generating a family planning report for the patient based on the probability that that the desired fertility outcome will result from each of the one or more retrieval cycles, and step 2208 may include providing the family planning report in any suitable interpretable format, such described in more detail herein. In some variations, the family planning report may include one or more indicators representing the probability of success for the desired fertility outcome, such as a first indicator for the predicted probability of success. Additionally, or alternatively, in some variations, the family planning report may include a second indicator comprising a plurality of steps of a fertility process (e.g., a timeline of an IVF cycle), and / or a plurality of third indicators each representing a predicted egg outcome for one of the plurality of steps. As described herein throughout, the method 2200 may additionally include using the family planning report to plan or adjust a fertility treatment for a patient, and / or to perform one or more associated treatment steps, such as, for example, to choose or perform a type of treatment, adjust or schedule a treatment start date and / or start the treatment, and / or choose to use and / or use PGT, donor eggs, and / or donor sperm.

[0145] FIG.23 depicts a flow diagram of another method for optimizing fertility treatment for a predictive family planning tool. The method 2300 may first include, at step 2302, predicting a success rate for a fertility outcome for a patient based on one or more models having received patient data. The fertility outcome may include live birth(s), such as two and / or three live births (and / or one live birth). The patient data may comprise one or more of: sperm parameters, age, BMI, hormone level (e.g., AMH level), infertility diagnosis, prior fertility outcomes (e.g., number of prior full term births), number of prior retrieval cycles, prior egg outcomes (e.g., number of prior eggs frozen), prior losses (e.g., number of prior pregnancies lost to miscarriage), treatment cycle number (e.g., retrieval number), transfer number from treatment cycle (e.g., from IVF retrieval cycle), storage status of one or more eggs (e.g., fresh or frozen), storage status of one or more embryos (e.g., fresh or frozen), and PGT status of one or more embryos. For example, the patient data may include at least the sperm parameters. The sperm parameters may include measurements (e.g., most- recent measurements, average of measurements over a predefined period (e.g., last month, last 3ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 months, last 6 months, last 9 months, last year) of sperm motility and / or sperm concentration. Accordingly, the method may include, prior to the step 2300, determining the sperm parameters from a sperm sample (e.g., from a patient^s partner or donor). In some variations, determining the sperm parameters may include determining sperm parameters from each of a first sperm sample and a second sperm sample, wherein the first and second sperm samples are from different people.

[0146] Next, at step 2304, the method 2300 may include providing an indicator for the success rate of the fertility outcome. In some variations, the method 2300 may be used to assess a potential IUI cycle for a patient. The indicator may be provided via a family planning report in the form of, for example, a verbal report (e.g., via a medical professional), a printed report, and / or an electronic report (e.g., a PDF, an email and / or via a graphical user interface (GUI) provided via a display). In some variations, the method 2300 may additionally include comparing the predicted success rate for the fertility outcome to another predicted success rate for a same fertility outcome that results from an IVF cycle to inform a treatment plan for a patient and / or to perform one or more treatment steps for a patient based on the predictions.

[0147] In some variations, one or more of the methods herein may be implemented together to predict one or more desired fertility outcomes. For example, the method 2200 may be implemented with the method 2100 so that predictions for a fertility outcome resulting from both of at least one IVF cycle (e.g., via the method 2100) and at least one IUI cycle (e.g., via the method 2200) may be provided to a patient for comparison.

[0148] In some variations, the methods described herein may be implemented as computer program instructions stored on a non-transitory computer-readable medium. When executed by one or more processors, these instructions cause the processors to perform the methods for optimizing fertility treatment as described herein.

[0149] Overall, the systems and methods described herein may provide practical improvements to both predictive modeling and assisted reproductive technologies. By generating individualized probability distributions for egg and embryo outcomes, the predictive family planning tool may enable earlier and more accurate scheduling of treatment, thereby reducing wasted cycles and preserving higher quality eggs and / or embryos. These predictions may allow physicians to adjust protocols, such as the number of retrievals, the use of preimplantation genetic testing, donor gametes, and / or dosing schedules, in a manner that limits unnecessary procedures, minimizesATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 patient exposure to invasive interventions and prolonged hormone therapy, and decreases use of clinical resources such as embryology lab time and operating room capacity. The cascaded model architecture, which may maintain probability distributions across developmental stages rather than collapsing them into a single outcome, may improve computational efficiency, support recalculations at various points during treatment when actual outcomes become available, and enhance overall predictive reliability. The PFP tool (e.g., one or more models thereof) may be retrained periodically or continuously on tens of thousands to hundreds of thousands of prior cycles, with drift-detection mechanisms to identify and correct model decay, further strengthening accuracy over time. Moreover, integration with EMRs and interactive GUIs may enable real-time report generation as patient inputs are adjusted, reducing manual data entry, improving consistency of counseling across providers, and standardizing the communication of clinical outcomes. Together, these features improve informed consent, continuity of care, and adherence to recommended protocols. Collectively, the systems and methods herein therefore improve fertility treatment outcomes such as faster time to pregnancy, higher likelihood of achieving desired family goals, reduced failed treatment cycles, and more efficient use of medical resources.

[0150] Throughout this application, the term ^about^ is used to indicate that a value includes the inherent variation of error for the device or the method being employed to determine the value, or the variation that exists among the samples being measured. Unless otherwise stated or otherwise evident from the context, the term ^about^ means within 10% above or below the reported numerical value (except where such number would exceed 100% of a possible value or go below 0%). When used in conjunction with a range or series of values, the term ^about^ applies to the endpoints of the range or each of the values enumerated in the series, unless otherwise indicated. As used in this application, the terms ^about^ and ^approximately^ are used as equivalents.

[0151] Additionally, it should be appreciated that ranges disclosed herein may be exemplary, and include all ranges and subranges therein.

[0152] While certain variations are described and illustrated herein, those of ordinary skill in the art will readily envision a variety of other means and / or structures for performing the function and / or obtaining the results and / or one or more of the advantages described herein, and each of such variations and / or modifications is deemed to be within the scope of the inventive variations described herein. More generally, those skilled in the art will readily appreciate that all parameters,ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and / or configurations will depend upon the specific application or applications for which the inventive teachings is / are used. Those skilled in the art will recognize or be able to ascertain using no more than routine experimentation, many equivalents to the specific inventive variations described herein. It is, therefore, to be understood that the foregoing variations are presented by way of example only and that, within the scope of the appended claims and equivalents thereto; inventive variations may be practiced otherwise than as specifically described and claimed. Inventive variations of the present disclosure are directed to each individual feature and / or method described herein. In addition, any combination of two or more such features and / or methods, if such features and / or methods are not mutually inconsistent, is included within the inventive scope of the present disclosure.

Claims

ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 CLAIMS What is claimed is:

1. A method for optimizing a fertility treatment for a patient, comprising: predicting a probability distribution for an egg outcome of a retrieval cycle for the patient based on a first model having received first patient data; predicting a success rate for a first fertility outcome for the patient based on a second model having received second patient data; and providing, via a report, a probability that a second, different fertility outcome will result from the retrieval cycle based on the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome.

2. The method of claim 1, wherein one or both of the first and second patient data comprise one or more of: age, BMI, infertility diagnosis, number of prior full-term births, number of prior retrieval cycles, number of embryos transfers from retrieval cycle, and preimplantation genetic test (PGT) status of one or more embryos.

3. The method of claim 1, wherein the first model is trained on a first training data set and the second model is trained on a second training data set.

4. The method of claim 3, wherein the first training data set comprises at least 1,000 prior- patient retrieval cycles, and the second training data set comprises at least 10,000 different prior- patient retrieval cycles.

5. The method of claim 1, wherein one or both of the first and second models comprise a regression model.

6. The method of claim 1, wherein the egg outcome comprises a number of eggs, a number of oocytes, a number of mature eggs, a number of post-mature eggs, a number of fertilized eggs, a number of zygotes, a number of embryos, a number of blastocysts, a number of usable blastocysts, aATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 number of euploid blastocysts, or a number of usable euploid blastocysts resulting from the retrieval cycle.

7. The method of claim 1, wherein the first fertility outcome comprises an embryo transfer.

8. The method of claim 1, wherein the second fertility outcome comprises a number of live births.

9. The method of claim 8, wherein the second fertility outcome comprises at least two live births.

10. The method of claim 1, wherein providing the probability that the second fertility outcome will result from the retrieval cycle comprises combining the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome.

11. The method of claim 1, wherein the first model is configured to predict an egg outcome attrition rate for the patient based on the first patient data, and wherein the predicted probability distribution for the egg outcome is further based on the predicted egg outcome attrition rate.

12. The method of claim 1 further comprising providing, via the report, an indicator of the predicted probability distribution for the egg outcome.

13. The method of claim 1, wherein the probability distribution for the egg outcome is a first probability distribution for a first egg outcome, and the probability that the second fertility outcome will result from the retrieval cycle is a first probability that the second fertility outcome will result from a first retrieval cycle, the method further comprising: predicting a second probability distribution for a second egg outcome of a retrieval cycle for the patient based on the first model; andATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 providing, via the report, a second probability that the second fertility outcome will result from a second retrieval cycle for the patient based on the second predicted probability distribution for the second egg outcome and the predicted success rate for the first fertility outcome.

14. The method of claim 13, wherein predicting the second probability distribution for the second egg outcome comprises predicting an attrition rate for the patient based on third patient data, and combining the predicted attrition rate with a predicted range for the second egg outcome.

15. The method of claim 14, wherein the predicted attrition rate comprises a rate of anti- mullerian hormone (AMH) decline, and wherein the third data comprises a future age of the patient.

16. The method of claim 1, wherein the egg outcome is a first predicted egg outcome, the method further comprising: predicting a second egg outcome based on a third model having received third patient data; and predicting the first egg outcome based on the second predicted egg outcome.

17. The method of claim 15, wherein predicting the second egg outcome comprises: determining a retention factor for the second predicted egg outcome; and adjusting the second predicted egg outcome based on the retention factor.

18. The method of claim 17, wherein the second predicted egg outcome comprises a number of eggs, a number of oocytes, a number of mature eggs, a number of fertilized eggs, or a number of zygotes, and wherein the first predicted egg outcome comprises a number of embryos, a number of blastocysts, a number of usable blastocysts, a number of euploid blastocysts, or a number of usable euploid blastocysts resulting from the second predicted egg outcome.

19. The method of claim 1 further comprising receiving, via a user, the second fertility outcome, wherein the second fertility outcome comprises a desired fertility outcome for the patient.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 20. The method of claim 1 further comprising determining a fertility treatment plan for the patient based on the predicted probability that the second fertility outcome will result from the retrieval cycle.

21. The method of claim 20, wherein determining the fertility treatment plan comprises one or more of: determining a type of fertility treatment for the patient, determining or adjusting a number of treatment cycles for the patient, determining or adjusting a start date for a treatment cycle for the patient, determining an egg freezing protocol for the patient, determining an embryo freezing protocol for the patient, determining a PGT protocol for the patient, and determining an egg type for the patient.

22. The method of claim 21, wherein determining the type of fertility treatment comprises electing an in vitro fertilization (IVF) process or an intrauterine insemination (IUI) process.

23. The method of claim 21, wherein determining or adjusting the number of treatment cycles comprises increasing or decreasing a number of treatment cycles.

24. The method of claim 21, wherein determining or adjusting the start date for the treatment cycle comprises advancing or delaying the start date.

25. The method of claim 21, wherein determining the egg freezing protocol comprises electing to freeze or to not freeze one or more eggs of the patient.

26. The method of claim 21, wherein determining the embryo freezing protocol comprises electing to freeze or to not freeze one or more embryos of the patient.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 27. The method of claim 21, wherein determining the PGT protocol comprises electing to test or to not test one or more embryos of the patient.

28. The method of claim 21, wherein determining the egg type comprises electing to use eggs of the patient or eggs of a donor during fertility treatment.

29. A method for optimizing a fertility treatment for a patient, comprising: predicting a probability for a fertility outcome of a retrieval cycle for the patient based on one or more models having received patient data, wherein the patient data comprises at least an infertility diagnosis and a transfer number from the retrieval cycle; and providing, via a report, an indicator of the predicted probability for the fertility outcome, wherein the fertility outcome comprises one or more of plurality of live births and a plurality of pregnancies resulting from the retrieval cycle.

30. A method for optimizing a fertility treatment for a patient, comprising: predicting a first egg outcome of a retrieval cycle for the patient based on a first model having received first patient data; predicting an egg outcome attrition rate for the patient based on a second model having received second patient data; determining a probability distribution for a second egg outcome of the retrieval cycle based on the first predicted egg outcome and the predicted egg outcome attrition rate, wherein the second egg outcome develops from the first egg outcome; predicting a success rate for a first fertility outcome for the patient based on a third model having received third patient data; predicting a probability for a second, different fertility outcome of the retrieval cycle for the patient based on the probability distribution for the second egg outcome and the success rate for the first fertility outcome; and providing, via a user interface, an indicator of the probability for the second fertility outcome of the retrieval cycle for the patient.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 31. A method for optimizing a fertility treatment for a patient, comprising: receiving patient data and a first predicted number of eggs resulting from a retrieval cycle for the patient; predicting, via one or more models, an embryo attrition rate for the patient based on the patient data; determining, via the one or more models, a probability distribution of number of embryos resulting from the retrieval cycle based on the predicted number of eggs and the predicted embryo attrition rate; and providing, via a user interface, a probability that at least two live births will result from the retrieval cycle based on the predicted probability distribution of number of embryos.

32. The method of claim 31, wherein determining the predicted probability distribution of number of embryos comprises combining the predicted egg outcome and the predicted embryo attrition rate.

33. The method of claim 31, wherein the eggs comprise oocytes or zygotes (2PNs).

34. The method of claim 31, wherein the embryos comprise one or both of euploid blastocysts and usable blastocysts.

35. A method for optimizing a fertility treatment for a patient, comprising: receiving patient data and a probability distribution of a number of embryos resulting from a retrieval cycle for the patient; predicting, via one or more models, an embryo transfer success rate for the patient based on the patient data; predicting, via the one or more models, a probability that at least two live births will result from the retrieval cycle for the patient based on the predicted probability distribution of number of embryos and the predicted embryo transfer success rate; and providing, via a user interface, an indicator of the probability that at least two live births will result from the retrieval cycle for the patient.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 36. The method of claim 35, wherein determining the probability that at least two live births will result from the retrieval cycle comprises combining the predicted probability distribution of number of embryos and the predicted embryo transfer success rate.

37. The method of claim 35, wherein the embryos comprise one or both of euploid blastocysts and usable blastocysts.

38. The method of claim 35, wherein the predicted embryo transfer success rate comprises a per- transfer euploid blastocyst success rate.

39. A method for graphically representing family planning predictions for a patient, comprising: receiving, at a processor, a desired fertility outcome for the patient and patient data comprising one or more of: age, BMI, AMH level, infertility diagnosis, number of prior births, number of prior retrieval cycles, number of prior eggs frozen, and number of prior losses; determining a probability that the desired fertility outcome will result from each of one or more retrieval cycles for the patient based on the patient data and one or more models; generating a family planning report for the patient based on the probability that that the desired fertility outcome will result from each of the one or more retrieval cycles; providing the family planning report via a graphical user interface (GUI), wherein the GUI comprises: a first indicator representing the probability that the desired fertility outcome will result from each of the one or more retrieval cycles; a second indicator comprising a plurality of steps of a fertility process; and a plurality of third indicators each representing a predicted egg outcome for one of the plurality of steps.

40. The method of claim 39, wherein one or more of the first, second, and third indicators comprise one or more of text, an image, a symbol, and a graph.

41. The method of claim 39, wherein the second indicator comprises a timeline.ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 42. The method of claim 39, wherein the predicted egg outcome comprises a number of eggs retrieved, a number of eggs frozen, a number of eggs thawed, a number of eggs fertilized, a number of blastocysts developed, or a number of euploids developed.

43. The method of claim 39, wherein the processor is configured to determine the probability that the desired fertility outcome will result from each of the one or more retrieval cycles, and to generate the family planning report.

44. The method of claim 39, wherein the processor is communicably coupled to a display configured to provide the GUI.

45. A method for optimizing a fertility treatment for a patient, comprising: predicting a success rate of a fertility outcome of a fertility treatment cycle for the patient based on one or more models having received patient data, wherein the patient data comprises one or more of age, infertility diagnosis, and sperm parameters; and providing, via a user interface, an indicator of the success rate of the outcome of the fertility treatment.

46. The method of claim 42, wherein the sperm parameters comprise one or more of sperm concentration and sperm motility.

47. The method of claim 42, wherein the fertility outcome comprises one or both of pregnancy and live birth.

48. The method of claim 42, wherein the fertility treatment comprises intrauterine insemination (IUI).

49. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform a method, the method comprising:ATTORNEYDOCKETNO.: ALFE-004 / 01WO 342409-2030 predicting a probability distribution for an egg outcome of a retrieval cycle for the patient based on a first model having received first patient data; predicting a success rate for a first fertility outcome for the patient based on a second model having received second patient data; and providing, via a report, a probability that a second, different fertility outcome will result from the retrieval cycle based on the predicted probability distribution for the egg outcome and the predicted success rate for the first fertility outcome.

50. The computer-readable medium of claim 49, wherein the report is provided via a graphical user interface (GUI) configured to receive at least a portion of one or both of the first or second patient data as adjustable input parameters.

Images