Medical image processing apparatus, and medical imaging apparatus
In medical examination of breast cancer, a lesion computer-aided detection is performed in real time and with high accuracy, and a burden on a medical worker is reduced. A medical image processing apparatus that processes a medical image includes: a detection unit configured to detect a lesion candidate region; a validity evaluation unit configured to evaluate validity of the lesion candidate region by using a normal tissue region corresponding to the detected lesion candidate region; and a display unit configured to determine display content to a user by using an evaluation result.
1. A medical image processing apparatus comprising:
an image processing unit configured to process a medical image acquired by a medical image imaging apparatus, wherein
the image processing unit includes a detection unit configured to detect a lesion candidate region from the medical image, and a validity evaluation unit configured to evaluate validity of a detection result of the detection unit,
the validity evaluation unit includes a normal tissue extraction unit configured to extract a normal tissue region from the medical image by using position information of the lesion candidate region detected by the detection unit, and a feature extraction unit configured to calculate a variation feature amount indicating a difference between features of the lesion candidate region and the normal tissue region, and evaluates validity of the detection result by using the variation feature amount calculated by the feature extraction unit.
2. The medical image processing apparatus according to claim 1 , wherein
the normal tissue extraction unit extracts a region adjacent to the lesion candidate region as the normal tissue region.
3. The medical image processing apparatus according to claim 1 , wherein
the normal tissue extraction unit extracts a region adjacent to the lesion candidate region in a horizontal direction as the normal tissue region.
4. The medical image processing apparatus according to claim 1 , wherein
the validity evaluation unit includes a layer structure analysis unit configured to analyze inclination in a horizontal direction with respect to the normal tissue region extracted by the normal tissue extraction unit, and
the normal tissue extraction unit corrects the normal tissue region by using the inclination acquired by the layer structure analysis unit, and passes the corrected normal tissue region to the feature extraction unit.
5. The medical image processing apparatus according to claim 1 , wherein
the validity evaluation unit includes a lower layer portion region extraction unit configured to extract regions below the lesion candidate region detected by the detection unit and the normal tissue region extracted by the normal tissue extraction unit, and a lower layer portion region feature extraction unit configured to calculate a lower layer portion variation feature amount indicating a difference in features between a lower layer of the lesion candidate region and a lower layer of the normal tissue region for the lower layer portion region extracted by the lower layer portion region extraction unit, and evaluates validity of the detection result by using the lower layer portion variation feature amount.
6. The medical image processing apparatus according to claim 1 , wherein
the feature extraction unit divides each of the lesion candidate region and the normal tissue region into small regions, calculates a set or a vector indicating a feature of each small region, and calculates the variation feature amount based on a similarity between a set or a vector calculated for the lesion candidate region and a set or a vector calculated for the normal tissue region.
7. The medical image processing apparatus according to claim 1 , wherein
the medical image is an ultrasonic image acquired by an ultrasonic imaging apparatus.
8. The medical image processing apparatus according to claim 7 , wherein
the feature extraction unit extracts features of the lesion candidate region and the normal tissue region by using a difference in the ultrasonic images between frames.
9. The medical image processing apparatus according to claim 1 , wherein
the validity evaluation unit further includes a lesion candidate region extraction unit configured to extract a lesion candidate region based on a detection result of the detection unit.
10. The medical image processing apparatus according to claim 9 , wherein
the lesion candidate region extraction unit compares detection results of medical images acquired temporally before and after, and changes an evaluation criterion for evaluating validity.
11. The medical image processing apparatus according to claim 1 , wherein
the feature extraction unit calculates the variation feature amount by using one or more of luminance of the medical image, a difference between time-series images, an edge feature, a texture feature, a frequency feature in combination.
12. The medical image processing apparatus according to claim 1 , wherein
the detection unit includes a convolutional neural network including a large number of convolution processes, and outputs position information and a certainty factor of a detection result of a detected lesion candidate region detected by the convolutional neural network.
13. The medical image processing apparatus according to claim 12 , wherein
the validity evaluation unit evaluates validity of the detection result by using the certainty factor output by the detection unit together with the variation feature amount.
14. The medical image processing apparatus according to claim 1 , further comprising:
a display control unit configured to display an evaluation result of the validity evaluation unit on a display device, wherein
the display control unit displays the evaluation result obtained by the validity evaluation unit together with the detection result obtained by the detection unit.
15. A medical imaging apparatus comprising:
an imaging unit configured to acquire a medical image of a subject; and
an image processing unit configured to process the medical image acquired by the imaging unit, wherein
the image processing unit includes a detection unit configured to detect a lesion candidate region from the medical image, and a validity evaluation unit configured to evaluate validity of a detection result of the detection unit,
the validity evaluation unit includes a normal tissue extraction unit configured to extract a normal tissue region from the medical image by using position information of the lesion candidate region detected by the detection unit, and a feature extraction unit configured to calculate a variation feature amount indicating a difference between features of the lesion candidate region and the normal tissue region, and evaluates validity of the detection result by using the variation feature amount calculated by the feature extraction unit.