Image analysis method and system for assessing bone fragility

1. A computer-implemented image analysis method, comprising: quantifying one or more features segmented and identified from a medical image including bone of a subject; extracting non-image data pertaining to the subject and pertinent to bone deterioration from one or more non-image data sources: extracting clinically relevant features from the non-image data pertaining to the subject; assessing bone fragility, comprising assessing the quantified one or more features segmented from the medical image and the clinically relevant features extracted from the non-image data with a trained machine learning model; and outputting one or more results of the assessing of the bone fragility: wherein quantifying the one or more features comprises determining any one or more of: vBMD (volumetric bone mineral density) of the bone, cortical porosity, transitional region vBMD, trabecular region vBMD, matrix mineralization level, marrow adiposity, average thickness of the compact cortex, trabecular tissue separation, trabecular density, and bone volume fraction.

2. A method as claimed in claim 1 , wherein the segmenting and identifying are implemented with a machine learning algorithm trained segmentation and identification model configured to segment and identify the features from the image.

3. A method as claimed in claim 1 , wherein the trained machine learning model comprises

(a) a disease classification model;

(b) a model trained using features extracted from patient data and labels or annotations indicating disease or non-disease;

(c) a deep learning neural network or other machine learning algorithms; and/or

(d) a model trained to diagnose and/or monitor one or more musculoskeletal diseases.

4. A method as claimed in claim 1 , further comprising (i) training the trained machine learning model, and/or (ii) updating the trained machine learning model with additional labelled data derived from new or newly analyzed subject data.

5. A method as claimed in claim 1 , wherein the results comprise

(i) one or more disease classifications;

(ii) one or more disease probabilities;

(iii) a bone fracture risk;

(iv) a fragility score;

(v) a porosity score;

(vi) a trabecular score;

(vii) a plot of fragility score evolving over time;

(viii) a plot of porosity score evolving over time;

(ix) a plot of trabecular score evolving over time;

(x) disease progression;

(xi) decrease in bone fragility; and/or

(xii) one or more fracture risk scores.

6. A non-transitory computer-readable medium comprising computer program code, wherein the computer program code comprises instructions configured, when executed by one or more computing devices, to implement the image analysis method of claim 1 .

7. A method as claimed in claim 1 , wherein the non-image data pertaining to the subject comprises non-image data pertinent to musculoskeletal disease, bone loss or microstructural deterioration.

8. A method as claimed in claim 1 , wherein the non-image data sources comprise video, voice, video of the subject while walking, and/or recorded voice from a subject interview data.

9. A method as claimed in claim 8 , comprising extracting one or more gait features.

10. A method as claimed in claim 1 , wherein extracting non-image data includes extracting structured and unstructured data from sources of structured and unstructured data, respectively.

11. A method as claimed in claim 1 , wherein the non-image data comprises unstructured data, and the method comprises (a) converting the unstructured data into structured data; (b) converting the unstructured data into structured data with a trained machine learning model; or (c) converting the unstructured data into structured data with a trained machine learning model that is a trained model of speech recognition or a trained model of optical character recognition.

12. An image analysis system, comprising: a feature quantifier configured to quantify one or more features segmented and identified from a medical image including bone of a subject; a non-image data processor configured to extract non-image data pertaining to the subject and pertinent to bone deterioration from one or more non-image data sources and to extract clinically relevant features from non-image data pertaining to the subject; a feature assessor configured to assess bone fragility, comprising assessing the quantified one or more features segmented from the medical image and the clinically relevant features extracted from the non-image data with a trained machine learning model; and an output configured to output one or more results of the assessing of the bone fragility: wherein quantifying the one or more features comprises determining any one or more of: vBMD (volumetric bone mineral density) of the bone, cortical porosity, transitional region vBMD, trabecular region vBMD, matrix mineralization level, marrow adiposity, average thickness of the compact cortex, trabecular tissue separation, trabecular density, and bone volume fraction.

13. A system as claimed in claim 12 , further comprising a segmenter and identifier configured to receive the image, segment one or more features from the image, and identify the features segmented from the image,

wherein the segmenter and identifier comprises a machine learning algorithm trained segmentation and identification model configured to segment and identify the features from the image.

14. A system as claimed in claim 12 , wherein the trained machine learning model comprises

(a) a disease classification model;

(b) a model trained using features extracted from patient data and labels or annotations indicating disease or non-disease;

(c) a deep learning neural network or other machine learning algorithms; and/or

(d) a model trained to diagnose and/or monitor one or more musculoskeletal diseases.

15. A system as claimed in claim 12 , further comprising a machine learning model trainer configured to update the trained machine learning model with additional labelled data derived from new or newly analyzed subject data.

16. A system as claimed in claim 12 , wherein the results comprise

(i) one or more disease classifications;

(ii) one or more disease probabilities;

(iii) a bone fracture risk;

(iv) a fragility score;

(v) a porosity score;

(vi) a trabecular score;

(vii) a plot of fragility score evolving over time;

(viii) a plot of porosity score evolving over time;

(ix) a plot of trabecular score evolving over time;

(x) disease progression;

(xi) decrease in bone fragility; and/or

(xii) one or more fracture risk scores.

17. A system as claimed in claim 12 , wherein the non-image data processor is configured to extract structured and unstructured data from sources of structured and unstructured data, respectively.

18. A system as claimed in claim 12 , wherein the non-image data comprises unstructured data, and the non-image data processor is configured (a) to convert the unstructured data into structured data; (b) to convert the unstructured data into structured data with a trained machine learning model; or (c) to convert the unstructured data into structured data with a trained machine learning model that is a trained model of speech recognition or a trained model of optical character recognition.

19. A system as claimed in claim 12 , wherein the non-image data pertaining to the subject comprises non-image data pertinent to musculoskeletal disease, bone loss or microstructural deterioration.

20. A system as claimed in claim 12 , wherein the non-image data sources comprise video, voice, video of the subject while walking, and/or recorded voice from a subject interview.

21. A system as claimed in claim 20 , wherein the non-image data processor is configured to extract one or more gait features.