Using a set of machine learning diagnostic models to determine a diagnosis based on a skin tone of a patient

Inventors

Swart, Elliot • Alivisatos, Elektra Efstratiou • Ferrante, Joseph • Asai, Elizabeth

Assignees

Digital Diagnostics Inc

Abstract

Systems and methods are disclosed herein for determining a diagnosis based on a base skin tone of a patient. In an embodiment, the system receives a base skin tone image of a patient, generates a calibrated base skin tone image by calibrating the base skin tone image using a reference calibration profile, and determines a base skin tone of the patient based on the calibrated base skin tone image. The system receives a concern image of a portion of the patient's skin, and selects a set of machine learning diagnostic models from a plurality of sets of candidate machine learning diagnostic models based on the base skin tone of the patient, each of the sets of candidate machine learning diagnostic models trained to receive the concern image and output a diagnosis of a condition of the patient.

Core Innovation

The invention provides a method for determining a diagnosis based on a base skin tone of a patient. A calibrated base skin tone image is generated using an image of the patient and a reference calibration profile, and a numerical representation of each pixel of the calibrated base skin tone image is determined. An aggregate representation is generated by performing a statistical operation based on the numerical representation of each pixel, and a point in a color space corresponding to the aggregate representation is identified.

The method maps the point in the color space into a discrete value in a classification system. A concern image of a portion of the patient's skin is received, and at least the concern image is inputting into at least one machine learning model. Information from which the diagnosis is derived is received as output from the at least one machine learning model.

The diagnosis is at least partially derived based on the base skin tone. Determining the diagnosis further comprises modifying the information from which the diagnosis is derived based on the base skin tone, and determining the diagnosis based on the modified information.

In dependent refinements, determining includes modifying calibration based on image suitability or quality, generating an adapted imaging profile from the reference calibration profile and the base skin tone and calibrating the concern image using the adapted imaging profile, and selecting a set of machine learning diagnostic models from candidate sets based on the patient’s base skin tone. Additional refinements include determining whether the concern image is diagnosable and outputting an indication when the concern image is not diagnosable, and determining the diagnosis by outputting a probability that the patient has a condition.

Claims Coverage

The partial content includes two independent claims: clm-00001 (a method) and clm-00009 (a computer program product). Across these independent claims, the coverage centers on a base-skin-tone-derived calibration and representation workflow, followed by machine-learning diagnosis using concern images with diagnosis information modified based on the base skin tone.

The independent claim coverage defines an end-to-end diagnostic determination framework where a calibrated base skin tone image is converted into a discrete classification via color-space mapping, and diagnosis derived from machine-learning analysis of a concern image is modified based on the patient’s base skin tone before the diagnosis is determined.

Stated Advantages

Documented Applications

No documented applications found