Systems and methods for processing electronic images to identify transplant donor-recipient matches
Systems and methods are described herein for processing electronic medical images to predict one or more donor recipients for a patient. For example, a digital medical image of the patient may be received, wherein the patient is in need of a transplant. A trained machine learning system may be determined. The digital medical image may be provided into the trained machine learning system, the trained machine learning system determining a patient embedding. Using the patient embedding, a subset of donor recipients may be determined. Based on the subset of donor recipients a recommendation of optimal donors may be determined.
1 . A computer-implemented method for processing electronic medical images to predict one or more donors for a subject, comprising:
receiving a digital medical image of the subject, wherein the subject is in need of a transplant;
determining a trained machine learning system;
providing the digital medical image into the trained machine learning system, the trained machine learning system determining a subject embedding;
determining, using the subject embedding, a subset of donors;
determining based on the subset of donors a recommendation of optimal donors; and
determining a second trained machine learning system, the second trained machine learning system being capable of determining a dietary, sleep, or exercise suggestion, wherein the second trained machine learning system receives as input the digital medical image of the subject, determines a lifestyle embedding, and determines, based on the lifestyle embedding, a dietary, sleep, or exercise suggestion for the subject.
2 . The method of claim 1 , wherein the transplant is a fecal matter transplant.
3 . The method of claim 1 , wherein the transplant is a liver transplant.
4 . The method of claim 1 , wherein a salient region detection module is applied to determine a saliency of each region within the received digital medical image, and non-salient image regions are excluded from processing by the trained machine learning system.
5 . The method of claim 1 further including:
receiving metadata corresponding to the subject, the metadata comprising: clinical data, genetic information, microbial composition, and/or life history data; and
providing the metadata into the trained machine learning system.
6 . The method of claim 1 , further including:
notifying, with a notification, donor collection entities of a need for donors with a given profile.
7 . The method of claim 6 , wherein the notification may include metadata of the subject in need, an indication of the donor profile needed that matches the recipient, and/or a request for the donor collection entities to start collecting donors that have a similar profile.
8 . A system for processing electronic medical images to predict one or more donors for a subject, the system comprising:
at least one memory storing instructions; and
at least one processor configured to execute the instructions to perform operations comprising:
receiving a digital medical image of the subject, wherein the subject is in need of a transplant;
determining a trained machine learning system;
providing the digital medical image into the trained machine learning system, the trained machine learning system determining a subject embedding;
determining, using the subject embedding, a subset of donors;
determining based on the subset of donors a recommendation of optimal donors; and
determining a second trained machine learning system, the second trained machine learning system being capable of determining a dietary, sleep, or exercise suggestion, wherein the second trained machine learning system receives as input the digital medical image of the subject, determines a lifestyle embedding, and determines, based on the lifestyle embedding, a dietary, sleep, or exercise suggestion for the subject.
9 . The system of claim 8 , wherein the transplant is a fecal matter transplant.
10 . The system of claim 8 , wherein the transplant is a liver transplant.
11 . The system of claim 8 , wherein a salient region detection module is applied to determine a saliency of each region within the received digital medical image, and non-salient image regions are excluded from processing by the trained machine learning system.
12 . The system of claim 8 further including:
receiving metadata corresponding to the subject, the metadata comprising: clinical data, genetic information, microbial composition, and/or life history data; and
providing the metadata into the trained machine learning system.
13 . The system of claim 8 further including:
notifying, with a notification, donor collection entities of a need for donors with a given profile.
14 . The system of claim 13 , wherein the notification may include metadata of the subject in need, an indication of the donor profile needed that matches the recipient, and/or a request for the donor collection entities to start collecting donors that have a similar profile.
15 . A non-transitory computer-readable medium storing instructions that, when executed by a processor, perform operations processing electronic medical images to predict one or more donors for a subject, the operations comprising:
receiving a digital medical image of the subject, wherein the subject is in need of a transplant;
determining a trained machine learning system;
providing the digital medical image into the trained machine learning system, the trained machine learning system determining a subject embedding;
determining, using the subject embedding, a subset of donors recipients; and
determining based on the subset of donors a recommendation of optimal donors; and
determining a second trained machine learning system, the second trained machine learning system being capable of determining a dietary, sleep, or exercise suggestion, wherein the second trained machine learning system receives as input the digital medical image of the subject, determines a lifestyle embedding, and determines, based on the lifestyle embedding, a dietary, sleep, or exercise suggestion for the subject.
16 . The non-transitory computer-readable medium of claim 15 , further including:
receiving metadata corresponding to the subject, the metadata comprising: clinical data, genetic information, microbial composition, and/or life history data; and
providing the metadata into the trained machine learning system.
17 . The non-transitory computer-readable medium of claim 15 , further including:
notifying, with a notification, donor collection entities of a need for donors with a given profile, wherein the notification may include metadata of the subject in need, an indication of the donor profile needed that matches the recipient, and/or a request for the donor collection entities to start collecting donors that have a similar profile.