Quantitative imaging biomarker for lung cancer

Patent metadata

Field	Value
Publication number	US-12548166-B2
Application number	US-202318176911-A
Country	US
Kind code	B2
Filing date	Mar 1, 2023
Priority date	Sep 2, 2020
Publication date	Feb 10, 2026
Grant date	Feb 10, 2026

Abstract

In one or more implementations, systems, methods and computer implemented processes are provided that are directed to a method of treating a subject with a lung tumor, the method comprising: obtaining computed tomography (CT) image slices of the subject, wherein the CT image slices comprise images of the lung tumor. In a further implementation, the systems, methods and computer implemented processes are directed to identifying a first CT image slice where the lung tumor has a largest diameter among the CT image slices; and determining intensity-skewness of the lung tumor on the first CT image slice. In a further implementation, the systems, methods and computer implemented processes are directed to treating the subject with surgery, chemotherapy and/or radiotherapy, if the intensity-skewness is no greater than −1.5.

First claim

The invention claimed is:

1 . A method of treating a subject with a lung tumor, the method comprising:
(a) obtaining computed tomography (CT) image slices of the subject, wherein the CT image slices comprise images of the lung tumor; (b) identifying a first CT image slice where the lung tumor has a largest diameter among the CT image slices; (c) determining intensity-skewness of the lung tumor on the first CT image slice; (d) classifying the subject to one of disease-free-survival (DFS)-associated histologic subgroup based on a comparison of the intensity-skewness to a threshold value; and (e) treating the subject with at least one of surgery, chemotherapy and/or radiotherapy, where the intensity-skewness is no greater than the threshold value.

2 . The method of claim 1 , wherein the intensity-skewness is:

Intensity_Skewness
=

1
N

⁢

∑

i
=
1

N

⁢

(

X
⁡
(
i
)

-

X
¯

)

3

(

1
N

⁢

∑

i
=
1

N

⁢

(

X
⁡
(
i
)

-

X
¯

)

2

)

3

wherein the first CT image slice N pixels: 1, . . . ith, . . . N, X(i) being an intensity of the ith pixel within the first CT image slice, X being a mean intensity of all pixels within the first CT image slice.

3 . The method of claim 1 , wherein the threshold value is accessed from a pre-trained machine learning model configured to output a value that corresponds to the demarcation between a first disease-free-survival (DFS)-associated histologic subgroup and a second disease-free-survival (DFS)-associated histologic subgroup-, wherein the pre-trained model is trained using a training set of CT images.

4 . A method for treating a subject with a lung tumor, the method comprising:
(a) obtaining computed tomography (CT) image slices of the subject, wherein the CT image slices comprise images of the lung tumor; (b) identifying a first CT image slice where the lung tumor has a largest diameter among the CT image slices; (c) determining intensity-skewness of the lung tumor on the first CT image slice; (d) determining that the lung tumor is a mid/poor disease-free-survival (DFS)-associated histologic subgroup, if the intensity-skewness is no greater than −1.5; and (e) treating the subject with at least one of surgery, chemotherapy and/or radiotherapy.

5 . The method of claim 4 , wherein intensity-skewness is:

Intensity_Skewness
=

1
N

⁢

∑

i
=
1

N

⁢

(

X
⁡
(
i
)

-

X
¯

)

3

(

1
N

⁢

∑

i
=
1

N

⁢

(

X
⁡
(
i
)

-

X
¯

)

2

)

3

wherein the first CT image slice has N pixels: 1, . . . ith, . . . N, X(i) being an intensity of the ith pixel within the first CT image slice, X being a mean intensity of all pixels within the first CT image slice.

6 . The method of claim 1 , wherein the lung tumor is lung adenocarcinoma.

7 . The method of claim 6 , wherein the lung adenocarcinoma is invasive adenocarcinoma.

8 . The method of claim 7 , wherein the

Assignees

• Univ Columbia

Inventors

• Lu Lin
• Zhao Binsheng
• Schwartz Lawrence H

Classifications

• G06T2207/30064

  Lung nodule · CPC title

• G06T2207/10081

  Computed x-ray tomography [CT] · CPC title

• G06T7/0012

  Biomedical image inspection · CPC title

• G16H30/20

  for handling medical images, e.g. DICOM, HL7 or PACS · CPC title

• G16H10/60

  for patient-specific data, e.g. for electronic patient records · CPC title